Progressive Education: On the Limits of Evaluation and the Development of Empowerment

Invited address presented at a conference, “Progressive Education: Reassessment, “jointly sponsored by Teachers College, Columbia University and Bank Street College, October 17-18, 1986.

Having demonstrated its durability, the progressive education movement is now being reassessed. These reappraisals will, inevitably, entail a search for relevant evaluation findings, now that it is fashionable and even obligatory to subject important educational innovations to systematic empirical evaluation. Thoughts about how children have been influenced by this mode of education are also bound to stimulate ideas about how its method and conceptual framework ought to be altered, given the rapid currents of social change. This article examines some of the complexities associated with designing and evaluating an educational program that is aimed at meeting the developmental needs of children. It asks whether the new methods of evaluation are applicable to forms of education that do not embrace traditionally defined academic objectives, and then initiates the complicated process of rethinking the goals for children within the framework of progressive education. It reviews the usefulness of the current commitment to evaluation and issues a call to revise and refine the child-development framework that undergirds progressive education.

EDUCATIONAL EVALUATION AND PROGRESSIVE EDUCATION

It is important at the outset to distinguish between the harsh assessment that I am about to make of our capability to conduct valid evaluations of educational programs and of the idea, in principle, of accountability. Clearly, it would be immensely valuable to be able to accurately chart the progress of children in school and thereby to gauge the effectiveness of teaching. However, I believe that the current state of evaluation methodology is too crude to even begin to provide the kind of data needed to render serious and dedicated educators accountable. It is easy, then, to vote with vehemence: for accountability, a resounding yes; but, for evaluation as it now stands as a means of achieving accountability, an equally resounding no.

As used in this discussion, the term evaluation refers to what has come to be known by the clumsy phrase “summative evaluation,” that is, evaluation that purports to indicate the degree to which a given educational program has achieved a specified set of outcomes among students. Thus, were we to adopt a psychological frame of reference in defining the task of evaluating education, as I think we should, then the job of the evaluator is to ascertain the degree to which the intended psychological changes in children to be effected by a given educational program have actually taken place, that is, to measure the psychological dimensions that the educator has been attempting to influence.

Yet the overriding problem that besets psychology is our inability to measure psychological processes with precision. This is the back-breaking problem that has threatened to defeat psychology, and has, at the very least, severely arrested its development. The gravity of the problem has evoked in psychologists extreme, even bizarre ways of coping with it, such as excluding from the purview of psychology all phenomena that cannot be directly observed and measured, or by defining important abstract concepts in terms of the crude methods with which they have sometimes arbitrarily and sometimes desperately decided to measure them.

In order to survive, psychologists have in one way or another learned to live with this nightmarish problem, just as many of us have learned to live with a severely limiting disability. As we are prone to say, some have even gone on to live productive and gratifying professional lives, but others, alas, have adopted coping mechanisms that are maladaptive. Most damaging from my viewpoint is the tendency to deny the disability by pretending to be able to perform exactly those acts they are manifestly incapable of performing. Among the most astonishing episodes of this pattern of denial has been the creation of an entire field of specialization, that of evaluation, that is called on to do precisely what psychologists, in their infancy, have yet to learn to do: to measure complex psychological processes. Whereas other psychologists have found creative ways to circumvent or deemphasize their disability, evaluation specialists are charged to measure, boldly and proudly, even though measurement is not really achievable—as if by designating them evaluators they will be exempt from the vicissitudes of measurement.

Evaluation specialists have responded to this dilemma by turning to those realms that are at least partly measurable, by overlooking the flaws in such efforts, and by assuming that these measurable domains can be studied in place of those that are directly relevant to the issue at hand. According to the argument that has evolved, the dimensions that can be measured are more real because they are measurable, and so they deserve to be regarded as more relevant; further, since all behavior is somehow interconnected, by studying what can be measured we are at least indirectly or partially assessing that which cannot be measured directly. Clearly, the only tenable solution to this thorny dilemma that evaluation specialists have been asked to live with is to use existing measures, and to assume that despite their deficiencies and their marginal relevance, assessing what can be measured is better than measuring nothing at all. What is more than a little preposterous, but not beyond understanding, is that all of these compromises and makeshift solutions to a state of affairs that is fundamentally unsatisfactory somehow bring about a feeling of certitude, if not downright virtue, among evaluators. An aura of definitiveness almost invariably surrounds the publication of evaluation findings. With so much denial, arbitrariness, and just plain wrongheadedness, it is no wonder that educators are wary of evaluation specialists.

When educational goals are defined as discrete and measurable entities, the work of evaluators is feasible, but when goals are defined in terms of a multiplicity of interrelated complex processes that may well be different for each individual, the rigidity and limitations inherent in evaluation technology stifle valid assessment.

The incompatability between evaluation methods and the goals of progressive education became plain to me when I first encountered the ideas and methods of progressive education. On coming to Bank Street as a young psychologist, and being introduced to Barbara Biber’s writings,1 I came to view all education, but especially progressive education, not as a means of implanting specific skills and pockets of knowledge, but as a form of profound, multifaceted psychological intervention in the lives of children. After having been trained to view children through peepholes defined by the atomistic one-variable-at-a-time research fashion of the day it was eye-opening to conceive of the classroom and the school as a psychological field, as a broad stage on which the full range of the drama and intricacy of psychological development is being enacted. The opportunity to deal with real children, each one obviously distinctive once looked at for more than a moment, and to bear in mind the full spectrum of their developmental needs and to project the sweep of development during a school career that spans more than a decade plunges the psychologist into a level of complexity and compelling reality that is at first staggering. It is a complexity that can only begin to be ordered and understood, at least from my experience, by viewing it through the lens of dynamic psychology.

Dynamic psychology not only provided a conceptual framework for explaining how child development was influenced by school experience. For me it served also as an analogue for understanding the relation of progressive education to traditional education. This relationship seemed to mirror the relation of dynamic psychology to the more dominant functionalist/behavioristic theoretical tradition of psychology. Just as dynamic psychology ran counter to the prevailing emphasis on studying observable behavior and conscious content by focusing on the role of the unconscious and inner life, so progressive education advocated a form of child-centered education that was out of step with the adult-centered education that had become one of our hallowed traditions. Calling for openness to experience, and pointing to the importance of personal choice and free expression, progressive education negated the traditional emphasis on achieving inner control by inhibition, and competence by adult modeling.

The interconnection between dynamic psychology and progressive education extended beyond the fact that they were both running upstream. In both there was the awareness of the strangulating effect of the overcontrol of feeling; the inhibiting effect and function of ritual, formality, and restraint; and the desirability of releasing creative energies that lay deep within. In both there was the understanding of the undermining role of tension, the deceptive appearance of orderliness and restraint and of our tendency to overvalue them, and the advantage of freeing people by reducing structure and allowing them to follow their own impulses and to become more aware of and reliant on their own resources. The issues of rigidity and freedom were central for both. Each responded to this altered framework by proposing a novel and distinctive mode of intervention that was more intricate than its predecessors and that required more training and sophistication on the part of the practitioner.

Both of these schools were regarded as far-fetched and iconoclastic, if not bizarre, and were often seen as undermining the very goals that their more traditionally minded co-professionals were struggling to accomplish, and as advocating a mode of professional implementation that was rebellious and could even become wild. Both introduced agendas that were infinitely more complex than what had preceded them. According to dynamic psychology, the family was an arena of much greater turmoil and psychic influence than had previously been imagined, and correspondingly, the progressive educator viewed the school as a potentiator of creative thinking, self-awareness, and inner strength far in excess of what was envisaged in the past.

I regarded both as forms of wizardry that were light years advanced in their theory and method. They had, in effect, transformed their professions from plodding mastodons to masterful agencies of insight and capability. Each had been exposed to ridicule and opprobrium by those most threatened by their ideas and, inevitably, each had been coopted. Yet it is impossible to stroll among the fields of psychology and education without coming on ideas and methods that originated in dynamic psychology and progressive education. Of course, this linkage between what I claim to be parallel developments in the fields of psychology and education is not a new idea. It has been intimated and often explicitly described in the writings of Susan Isaacs, Barbara Biber, Lois Murphy and H. Ladd, and Lawrence Kubie,2 among others, who, in their own creative ways, extracted, integrated, and recast the basic tenets of dynamic psychology into a framework that guided educational practitioners, especially those working within a progressive framework.

For the sake of accuracy, it should be noted that elements of dynamic psychology were selectively and variably adapted, and that there were sectors of progressive education that were more philosophically than psychologically driven, and others that were moved less by ideological or psychological guideposts and more by the impulse to be open and experimental, and there were still others that were mainly just plain rebellious. Also contributing to the emergence of progressive education was a keen disappointment with the aridity of the academy that frequently evoked in this alternate form of education strong currents of anti-intellectualism. Some strands of progressive education were dominated by a surge toward freedom and exploration; their vision was directed outward toward the challenge of the future rather than inward and to what had been learned in the past. Enthralled with the instrumental potentiality and natural beauty of the environment, they had neither time nor appetite for the introspective analysis of the psyche that some regarded as the mambo-jumbo of twentieth-century prophets and witch doctors.

I have taken the time to portray the wide-ranging, heterogeneous quality of the world of progressive education in order to emphasize that this discussion of the relation between evaluation and progressive education deals exclusively only with that segment of progressive education which, either explicitly or intuitively, has incorporated the insights of dynamic psychology. Progressive educators of this persuasion might well view themselves as applied dynamic psychologists, as professionals who function in school settings to intervene in support of the psychological well-being and growth of children. In such cases, the aims of education, either explicitly or implicitly, become redefined and expanded to include an array of themes that are best summarized under the rubric of the promotion of ego functioning. The boundaries of this wing of progressive education clearly overlap with what later came to be known as an orthopsychiatric view of education.

I speak of the parallels between and even the convergence of dynamic psychology and progressive education as a way of emphasizing how unsuited the conventional methods of evaluation are to assessing the impact of progressive education—in the same way that the tenets of dynamic psychology have been resistive to assessment and verification. Dynamic psychology deals with fluid rather than discrete and static facets of behavior. It explores the inner life and hidden motives; it recognizes the unity of opposites in accounting for the motivation of behavior. It is primarily concerned with integrative levels of psychological functioning, with overarching directions and goals, and with styles of thinking and conflict resolution. Less concrete and atomistic in its analytic mode, and not at all bound to deal exclusively with observables that can be quantified, it presents a territory that is unfriendly to measurementminded psychologists.

In similar fashion, progressive education is more receptive to subjective phenomena, and embraces more molar units of analysis. It recognizes the importance of global or atmospheric variables that cannot be easily quantified, factors such as the emotional climate of a classroom and the personal meaning that is associated with a particular cognitive task. Because progressive educators know that children learn more deeply when they are fully abandoned to and immersed in the task of learning and when they are allowed to function autonomously and to express their individuality, they are less regimenting and coercive, more willing to try multiple routes to a goal, and less rigidly concerned with achieving the same goals for all children.

The learning tasks in traditional classrooms are more segmented and explicit, and expectations of student achievement are more uniform. The teacher asserts greater control over what each child is exposed to and is more explicit and concrete about what is to be learned. Such programs by their very nature are more amenable to educational evaluation. Classrooms in progressive schools are seen more as vehicles for supporting complex integrations and new initiatives than as shops for the production of particular outputs. Not only is it more difficult to describe and measure such forms of educational/developmental change because of their fluidity and complexity, but different patterns are assumed for different children, so that the same measuring rod can seldom be applied to all children.

Another way to contrast our ability to measure the more fluid goals of progressive education (defined in terms of the tenets of dynamic developmental psychology) and the more explicit, mechanistic goals of traditional education is to invoke a distinction made by Florence Goodenough almost a half-century ago.3 She noted that some tests consist of signs and others of samples. Implicit in this distinction was the greater technical difficulty associated with measurements based on signs as compared with those based on samples. When a test is constructed to sample a specifiable, circumscribed domain, as in the case of spelling, arithmetic, and other forms of conventional educational achievement, measurement is on much sounder ground than when we attempt to measure complex forms of psychological organization by detecting signs of their presence.

It is not merely that the distinctive and more numerous and comprehensive goals of the progressive educator are more difficult to measure. Clearly, the validity coefficients of tests of self-reliance, resourcefulness, imaginativeness, and ego functioning- to name a handful of dimensions that come immediately to mind—would be unacceptably low, thus dooming attempts to measure the effectiveness of progressive education to the crudest level of measurement even if tests and measures of the relevant variable existed. More important, the use of tests implies a rigidity and uniformity to the operation and functioning of the classroom, and a compartmentalization of its concerns to discrete units, that is incompatible with the ethos and operational style of progressive education.

The fluid, dynamic, process-oriented character of the progressive education classroom notwithstanding, we might nevertheless maintain that it would be of interest to administer a series of standardized achievement tests, in passing, just to see how children in such classrooms function on such instruments. The problem lies in the exaggerated importance, the false validity that is imputed to evaluation data, the aura of definitiveness that they cast (even when evaluators are modest in their claims), and the tangled web of comparisons and wrong inferences they invite. Let us look at a specific body of recent research in order to better understand what is at stake.

On the basis of a series of studies, researchers have recently concluded that the more time given to explicit instruction in a particular curriculum area- “time on task” is the official jargon used to denote this variable—the better will achievement be.4 These findings have been touted as pointing to the solution of our educational problems. Where children are not learning effectively, we need only spend more time teaching them. In effect, this recent pocket of research findings demonstrates and documents how data that bear on the issue of evaluation can be used to drive the curriculum.

These findings are at once self-evident and mischievously deceptive. On the one hand, it is perfectly clear that children need to spend some time studying or at least being exposed to material in order to learn it, and that frequently the opportunity to spend more time enhances the likelihood that learning will take place. From one standpoint, then, such findings are hardly surprising. One may even wonder why so self-evident a proposition had to be studied.

At the same time, these findings mislead by oversimplifying. That students who spend more time dealing with particular material learn more may be a reflection of the fact that they are more interested and more receptive to learning, are therefore more willing to spend time working on a task to be learned, and in turn, likely to learn more. For such children, the greater time on task may be a reflection or a consequence of their greater learning ability rather than the cause of it. In addition, some teachers may spend less time with a particular subject matter, not because they are indolent or fail to recognize that it would be useful for their children to have greater exposure to the material, but because they have found that their students lack the motivation or attention span for investing more time. In many instances, then, lessened time on task may be more of a symptom of learning difficulties than a cause. When this is the case, using such evaluation data to induce teachers to spend more time on task may run counter to the best information regarding the limited tolerance of a particular student body for dealing with such material. In other cases, the lesser time on task may reflect the priority system of a given teacher or school. In the competition for allocation of classroom time, one particular activity warrants less attention than others. Further, most efforts to teach effectively can be improved more by changing the method of teaching, by rethinking the strategies that have been used, rather than merely applying more time to a method that may have been defective in the first place. Yet the data on time-on-task studies have been heralded as providing new insights into the dilemma of improving the level of public education, and educators have been urged, and in some cases ordered, to devote more time to teaching particular subject matter. In this fashion, research data, no matter how senseless they may be, are used to drive the curriculum.

What is likely to be the impact of the new interest in systematic application of the findings of the technology of evaluation? The answer to this question is by now familiar. It will lead to patterns of teaching to the test, as if tests were the repository and embodiment of ultimate educational wisdom rather than hurriedly assembled samples of measurable aspects of material to be learned. More disturbing, it will also, in some cases, lead to such desperate measures as the falsification of test data, in order to forestall unfavorable and/or unfair assessments of educational effectiveness. Even in those schools where recently obtained evaluation data are officially recognized to be imperfect, tangential, and only marginally relevant, their empirical, scientific, quantitative, documented quality marks them as a body of data that might be seized on by officials who are new to the situation (how could they resist the temptation?)—critics, adversaries, or mischief-makers -and exploited by them to achieve their ends. Such a state of affairs quite understandably leaves teachers with a sense of discomfort and anxiety about the very existence of such data, and behooves the practical, more self-protective members of their lot to pay attention to the implications of testing.

Evaluation data become a constraining force to the more spirited and creative teacher, a force so dampening that such teachers may become discouraged from continuing to teach. Evaluation data may serve as a beacon to teachers who lack a sense of purpose and need more structured guidelines for engaging in their day-to- day work. Thus, evaluation methods are selectively helpful and disruptive; but their overall effect appears to be to drive the quality of education to the lowest common denominator. In sum, I believe that educational evaluation is an intrusion that produces a distorted image of what school is about to students, teachers, parents, and the community, and to most schools, but especially to progressive schools.

In those rare schools where the educational goals are precisely matched by the content of achievement tests used to evaluate education, educational evaluation would be valid and would prove immensely valuable to the administration of schools. Where there is a disparity between the content of tests and the goals of the educators, however, the process of educational evaluation is more likely to be distracting and undermining than illuminating. As tests are made more efficient to use by new technology, they may well become an irresistible force in a school community. Their use feeds on itself. As more time and resources are given to educational evaluation, testing procedures become an increasingly more potent force, one that erodes the autonomy and quality of teaching. Further, as the curriculum becomes driven by educational evaluation, the teaching profession becomes less challenging, and is likely to attract less competent people.

To summarize: I believe that out current ability to assess educational effectiveness is too imperfect to be generally useful, and that our measurement technology is applicable and informative to only a comparatively small region of educational objectives (that are already being accorded disproportionately more attention and importance than they deserve by virtue of this fact). It will require much more time and vastly greater resources than are now available to begin to be able to comprehensively assess the impact of schooling, were we technically able to perform such a feat. We tend to overestimate the validity of current measurement capability and to correspondingly underestimate the time and complexity—and, therefore, the ultimate cost—of valid and comprehensive educational evaluation. When the day comes that we will have the capability to conduct such comprehensive evaluations with validity, it will be a millennium of sorts. The school of the future will be a ten-story building in which the ground floor will be given over to classrooms and the remaining nine-tenths of the building will be devoted to the analysis and reporting of evaluation data. This hyperbolic description is offered in order to emphasize the immense complexity entailed in comprehensive educational evaluation when and if we develop the methods to conduct it properly. In the meantime, educational evaluation is extraordinarily narrow in its scope and imperfect in its measurement. It is singularly inappropriate for assessing educational programs with multifaceted goals. It is even less appropriate where education is highly individualized, thereby introducing the added burden of using idiographic rather than nomothetic modes of measurement.

The serious limitations associated with the low validity of the findings of educational evaluation (and when I speak of low validity I refer to both inaccuracy in measurement and, to a greater extent, irrelevance of measurement) and the pressures that evaluation programs implicitly bring to bear on teachers, pressures that impel them to use methods and to teach content that may not fit in with their vision of what would most benefit the children in their class (or worse, materials that discourage them from even having a vision of the educational process that differs from that which is implicitly stated in evaluation programs), thereby impair the depth, coherence, and power of their work.

In light of these considerations, I have for some time urged that we choose an alternative path to evaluation.5 Since we now lack the capability to measure with precision the most relevant attributes in each child that are affected by school experience, that is, since we are unable to accurately assess the main lines of school influence, we should instead direct our evaluation efforts to assessing the quality of the school environment. It is much more difficult to pinpoint and gauge how each child has been affected by his or her school experience than it is to describe in detail the character of the school environment to which each child has been exposed, and then to estimate the quality of school impact.

What I am proposing is hardly radical; it is how our most telling but less formal evaluations are now conducted. If we are considering whether to enroll a child in one school or another, or whether to live in one neighborhood or another, or whether to vacation in one city or another, we do not consult data regarding the psychological impact of such experiences—not because we are not interested in such information but because we know that they will not be forthcoming in reliable fashion. Even if such data were available, we would have the good sense to recognize that they would be too fragmentary and inaccurate to serve our purposes. Instead, we are likely to base our decision on as thorough an assessment as can be obtained of what these alternative stimulus situations have to offer and then to make inferences regarding their potential impact on us. For example, if we intend to travel, we will consider the degree to which each site offers the right balance of leisure, comfort, and stimulation. In sizing up a summer camp, we think about whether our child will have nice friends and be cared for by nurturant and competent caretakers, and whether there will be facilities that will engage and stimulate the child in ways that will lead to positive psychological growth. Similarly, in a school, we look at the child group, teaching staff, and facilities, and in making our assessments, think about their potential long-term impact. I argue that this is the manner in which important evaluations are now made and should be made—until we develop the technical capacity to comprehensively assess the psychological impact of various environments.

Having advocated this alternative route to evaluation ten years ago, I regret not having had the time and motivation to provide further support for the implementation of this recommendation by developing a framework for the comprehensive assessment of school environments. A set of guidelines for systematically gauging school environments is needed to facilitate such work. However, I have never doubted that the general path I have recommended, and paradoxically the one that is already most heavily traversed without being officially acknowledged (because it is conducted informally and without the trappings and rituals of social science), is the one that should be taken by most official and weighty evaluations. The study of the actual psychological impact of educational environments should, of course, continue—but it should proceed under the aegis of research and should be recognized as being in its early, formative stages. Its findings should never be presented as constituting definitive evaluation data.

SEARCHING FOR THE IMPACT OF PROGRESSIVE EDUCATION ON CHILDREN

Having taken time to question the usefulness of evaluation methods to progressive education, I want now to discuss some ideas about the impact of progressive education on children. In effect, I will speculate about what we might find if it were possible to subject progressive education to thorough and valid summative evaluation. Paradoxically, I will call on data based on methods related to, but different from, those I have taken such pains to discredit. I refer to a body of findings obtained several decades ago by Pat Minuchin, Barbara Biber, Edna Shapiro, and myself. along with many other colleagues in the Bank Street Research Division, in a study entitled The Psychological Impact of School Experience.6 In that study we attempted to identify the nature of psychological differences that might be found between children attending what we then called modern (but what were in many important respects progressive) schools, and other nine-year-old children of comparable upper-middle-class family backgrounds who had attended traditional schools.

It may seem outrageous to now call into play a set of findings obtained from methods that I have just criticized, but, as you will see, we regarded our findings as tentative, and left the study acutely aware of the limitations and pitfalls of comprehensively assessing the psychological impact of school experience. Perhaps the most distinctive feature of our study is that we made a serious attempt to be comprehensive in our assessments of children and appropriate in our group comparisons. Although the study was limited to slightly over one hundred children, it took over a decade to plan the work fully and to gather and analyze the data. Even though we made individual assessments of each child on repeated occasions using a wide variety of measures, our study raised more questions than it answered, and left many basic questions about the differential influence of progressive schools still in the dark. At the same time it yielded some informative results, and others that were suggestive. It should be emphasized, however, that such an intensive and costly approach to the problem of evaluation could never be adopted by small evaluation units working with necessarily modest budgets and charged with routinely assessing huge numbers of cases.

Now let me turn to a pair of findings that I wish to discuss. One has to do with the fact that the progressive school movement appears to have brought out in its children, at that point in time, patterns and perspectives that were comparatively rare but are now visible in most children. In effect, the progressive movement was a harbinger, a kind of California of American childhood—a place where things seem to have happened first. The children whom we studied from progressive schools a generation ago showed themselves to be more open, less influenced by social sex-role expectations, more accustomed to making choices, more closely tied to their peer group, and seemingly possessing an interpersonal stance when facing adults that enabled them to hold their ground and to express themselves more easily and directly. In effect, the progressive education movement was concerned with inducing values in children and relating to children in ways that presaged patterns that later became the prevailing mode in the larger population. Whether the progressive education movement provided the soil for such changes, or functioned as a conduit or outlet for them by simply serving as the kind of school that was preferred by families who were among the earliest to embrace such values, is difficult to disentangle. (Of course, these are not mutually exclusive explanations of what may have transpired.)

I cite this observation, not because trend-setting power is in and of itself particularly admirable, but because, in this case, it seems to attest to the authenticity of the movement’s goals, as if there were an inevitability to them that bespeaks their truthfulness. It suggests, further, that the movement has had far greater impact than is usually acknowledged. It seems to have articulated goals and set into motion methods for achieving them that, while not immediately and entirely accepted, became the cornerstone of new values and patterns once they were promoted by more trusted and less threatening agents in our society. Of course, these values and patterns were frequently modified in ways that made them more palatable but that deviated significantly from the original intent.

In the portrait of distinctive features associated with children attending progressive-schools that emerged from our study was a pattern-not robust, but provocative. There were indications among our data that children from progressive schools took themselves more seriously as autonomous, willful creatures, and therefore viewed themselves as having the potential to have a serious (including also negative) impact on their environment. This observation was derived from the children’s responses to some questions that were designed to assess moral patterns of development and issues pertaining to impulse control. Among these questions were two of special interest: What is the worst crime? and What is the worst thing that a child in your class could do? When asked about the worst crime, the modal response of children in both progressive and traditional schools consisted of references to acts of murder or of killing someone. In sharp contrast, when asked about the worst thing a child in their class was capable of doing, children from traditional schools usually referred to a harmless prank. On the other hand, a sizable number of children from progressive schools described children, as they had adults, as being capable of the act of murder. This differential pattern of responding was surprising in light of the more explicitly humanistic ethos of the progressive school environment and the rather clear-cut tendency that we had found for children from progressive schools to be more clearly identified with childhood—to embrace and value the stage of childhood more and to be more inclined to view the world from a child’s perspective. Here I will digress to discuss these related findings in greater detail.

Some of the most interesting work in our study was aimed at demonstrating that the children whose modes of psychological functioning we were contrasting—those from modern or progressive schools and those from traditional schools—assigned different value to their status as children. We were interested in this dimension (one that I believe has never been previously studied nor studied since) because of our conviction that one of the main features of progressive education was that it accorded children and childhood respect. This was accomplished in part by creating an environmental framework that expressed and reflected children’s values, interests, and concerns. It encouraged them to build on what they already knew and had experienced, and to follow their paths of curiosity. In effect, an environment was created that was empowering.

In marked contrast, traditional schools were designed to remind children—by means of their decor, their curriculum and expectations, and their emotional—climate that school was there to help, if not require, children to leave behind their childish ways. School was the main train, really the last train, to adulthood. We reasoned, therefore, that it would be interesting to be able to demonstrate that children from progressive schools would show a greater affinity for the status of being a child, and a greater likelihood to view the world from the perspective of a child, and that children from traditional schools would assign greater value and urgency to achieving adult status.

We succeeded in demonstrating that this pattern existed, and I think that we regarded the data bearing on this issue as the most revealing and methodologically successful accomplishments of our study. We believed that it was enabling for children to feel comfortable and potent within the framework of their own developmental epoch, and, conversely, that it was undermining to the child’s sense of self to be relentlessly reminded that adulthood was the promised land and that they could be rescued from their adrift state only by paddling to the distant shores of adulthood.

In light of these considerations, it was somewhat surprising to find that the children from the benign, child-centered world of progressive education were more likely to view themselves, in effect, as capable of committing murder, and that it was the children from the adult-centered educational experience who viewed themselves as being little more than pranksters. However, this seemingly anomalous finding could be explained by remembering that one of the main reasons for helping children to remain comfortable in the world of childhood, to use the appetites, interests, and resources available to them as children in their daily explorations and learning, was the expectation that is empowering to live within such a framework.

I believed that this difference in how children described the worst thing that they were capable of reflected the fact that children from traditional backgrounds did not yet feel entitled or competent to commit acts that mattered. They were more apt to view the process of growing up as a sort of apprenticeship to adulthood, as a gradual passage into the world of adult prerogatives and adult power. Until they reached that privileged position, children saw themselves as still harnessed and unready, and therefore less potent. Included in this more modest self-appraisal that traditional celebrations of the rites of passage have helped to express was the idea that they were not yet capable of having the destructive capacity of adults.

In marked contrast, children from progressive schools, although more embedded in the world of childhood and less likely to invoke adult values in their choices and assessments, nevertheless regarded themselves as potent in and of themselves. They did not appear to view their child status solely in terms of their distance from adulthood. As a result, when it came to judging children’s capacity for destructiveness, they saw children as not so different from adults. They were more enmeshed in the child’s world, but this world was not primarily characterized by impotence, an impotence that could only be overcome by entering the world of adults. In effect, children from progressive schools were more inclined to regard themselves as empowered, as having teeth.

I have chosen to talk about this issue for three reasons. First, it demonstrates how complex and subtle are the phenomena of school effects—especially when we take into account the distinctive agendas and goals of progressive education. The domain of empowerment, while clearly of great theoretical usefulness, is not a dimension of psychological functioning that is well understood, nor is it easily observable. It entails developmental patterns and psychological issues that are particularly resistive to systematic measurement. It is an example of the vast complexity and scope of assessment that would be needed to begin to comprehensively assess the psychological impact of school experience, especially progressive school experience.

Second, it is another illustration of how the progressive schools of yesterday frequently anticipated the psychological trends of today. The psychological development of today’s children is an issue that has captured my interest. In a recent exploration of how children and childhood are changing, based on retrospective perceptions of intergenerational change obtained from several hundred experienced professionals, I have to come to believe that children today display an increased sense of empowerment, similar to but not identical with that which was observed among children from progressive schools in the previous generation.7 In attempting to track the main axes of psychological change in today’s children, I am struck by the salience of this difficult-to-pin-down attribute.

However, the changing patterns of empowerment that are visible in today’s children are not exactly like with those previously found among children in progressive schools. The enhanced empowerment of children from progressive schools in the previous generation seemed to be derived from their greater sense of competence, from their more autonomous learning experiences and the reassurance they gained from a nurturant and supportive learning environment. On the other hand, the manifestations of a sense of heightened empowerment that seem to be so pervasive in today’s children appear to have been arrived at by a different route. They seem to result from the extra support that technology provides, from the decline of traditionalism and the atrophy of prohibitions and strictures, and—perhaps more significant—from dramatic changes that have taken place in the manner in which we relate to authority. There is also an increased independence that not infrequently originates in parental neglect and feelings of isolation. Thus there appear to be important differences in these parallel trends that point to the fuzziness of the concept of empowerment. This concept is no doubt a shorthand description for a complex set of ideas that need to be better sorted out and understood. As this construct becomes more differentiated, we will become increasingly aware of distinctions among its various manifestations that are important to register.

My third reason for discussing this finding follows on the second. It reminds me that the need for deepening and clarifying our understanding of theoretical constructs is more general. Our conceptual framework for describing the developmental needs of children that can be influenced by school experience is powerful but incomplete. It needs to undergo further elaboration and differentiation, and it also needs to be reexamined in light of the magnitude of recent technological and social change.

During the early days of refashioning the classroom and educational strategy to fit what was known about child development we were able to identify particular dimensions of child life and child development that it was important to influence by means of school reform. At that time, our inclination was to concentrate on effecting change in those aspects of psychological functioning that were manifestly far from the optimum. It was as if we had suddenly found ourselves in a dark and musty building in which all of the windows were sealed and the atmosphere was stifling. Our initial, valid response was to open the windows as wide as possible and let in some air. We assumed, for the most part correctly, that children needed more, not less freedom; more, not less empowerment; more, not less self-assertion; less, not more feelings of guilt; more, not less opportunity to express feeling. Now that we have opened the windows, however, it has become positively drafty in some spots while others remain stifling. It is clear that we need to develop a more refined and sophisticated concept of optimal ventilation. Much of our original thinking was influenced by Freud’s powerful explanation of latency phenomena and his delineation of the massive role of anxiety, emotional conflict, and the defensive use of repression during childhood.8 In order to advance beyond this ground-breaking work, we need to refine and revise the conceptual framework that contributes to our understanding of how schooling affects child development.

In considering the concept of empowerment, for example, we need to press for identifying its defining properties. To what extent is empowerment characterized by feelings of entitlement, and by greater self-assertiveness? What other elements comprise this concept? Once we have achieved a clearer and more comprehensive definition of the construct, we will need to examine its ramifications for children at different developmental stages. What are the corollary levels of social responsibility, self-understanding, and sense of purpose that are associated with the development of a sense of being empowered, and how is the quality of empowerment affected by variations in these associated characteristics? What are the differences between feelings of empowerment that are associated with a sense of purpose and highly developed social responsibility and those that are not?

In a similar vein, we need to subject to closer examination our ideas about other core themes that were central to the progressive movement, ideas about relations with authority, individuality, freedom, the role of guilt, and the expression of affect. In the course of refining this conceptual framework we need also to bear in mind that the cross-section of children with which educators now work, and with which most children interact, is much broader and quite different from the children with whom the pioneers in our field dealt, different with respect to their family constellations and background and their childhood experiences. In raising questions about the universal applicability of some facets of dynamic developmental psychology I am not at all suggesting that we overturn or abandon them, but that we become more alert to the possibility that some of the propositions and variables that we have borne in mind in dealing with children may not be as relevant when applied to different child populations, or may show different patterns of functional relationship among them.

Equally important is the need to take into account the changing world of childhood. While I cherish this occasion for arriving at a fuller appreciation of the major contribution of progressive education to our thinking about children and our ways of educating them, and to honor and celebrate these past accomplishments, it is now time to reexamine and extend those very fertile lines of thinking in the light of new knowledge and the extraordinarily rapid currents of social change.