Remarks Prepared for “The 2000 Sociology of Education Association Conference,"  The Asilomar Conference Center, Pacific Grove, CA. February 25-27, 2000. 

Ronald G Corwin, Professor Emeritus

Department of Sociology, Ohio State University

I appreciate this opportunity to be here with you to reflect on where Sociology of Education has been, and to speculate about where it might go.   Media hype suggests that the new millennium will usher in spectacular changes in the society– which presumably foreshadow dramatic changes in our specialty as well.  Perhaps.  But a more sober view of our future is that it will be just like the present – only longer!  Okay, so that’s probably not accurate either.   Perhaps a developmental perspective is the most fitting.   As I reflect on our history, it seems to me that our specialty is nearing the end of its adolescence and entering a mature phase of development. 

But the question I want to address here is, will sociology of education ever grow up?  I maintain that depends on the willingness of those of us in this room to recognize and, more important, to apply some basic lessons that we should have learned during the past century.
 

Historical Highlights

Before getting to those lessons, I want to remind us about three crucial aspects of our history, which have formed our character, and in some ways marred our character.

First, we can’t afford to forget that sociology of education spent its childhood at the margins of both sociology and education.  So-called “educational sociology ” was conceived from a mixed marriage—a marriage between amelioration and empirical sociology.  While Durkheim was pleading for research on the social “facts” of education, many American sociologists were preoccupied with improving public schools.  Lester Ward, for example, regarded education as the “mainspring of all progress,” and shared the widespread opinion that education would cure most of society’s ills.  Many of us believed him.  And, as a bonus, he told us how to do it.  But I also have to mention Willard Waller here, because I admire him as one of the first sociologists in this country to take an objective sociological approach to schools. 

Of course the discipline of sociology as a whole was bent on amelioration.  Albion Small said that Sociology “was born of the modern ardor to improve society.”   But if that was of sociology, it was especially true of educational sociology.

So?  What of it?  That brings me to a second feature of our history—which is this: the early years were not very productive.  Estranged from the discipline, educational sociology took its cues from practitioners.  Nourished within a mixed marriage, we grew tall—but we remained scrawny.  By the Great Depression, over 25 textbooks in “Educational Sociology” were being used in more than 16 universities.  But most educational sociologists were trained in education and used substandard methodologies and a-theoretical approaches.  And, they were preoccupied with narrow, pragmatic, short-term problems.  It is revealing that, writing as recently as 1962, Andrew Halpin questioned whether our findings were even worth disseminating.

The third aspect of our history that I want to mention concerns the indelible scars left by events during the 1960’s – when our specialty entered a turbulent period of adolescence.  With adolescence, we developed a heightened identity crisis.  Like the proverbial fatherless son struggling to escape his past, we changed our name to distance ourselves from our pedestrian path.  Henceforth, we would be “sociology of education”.  And we made invidious distinctions between colleagues trained in sociology and appointed to sociology departments and those who were not. 

         Also, a cocky arrogance made our adolescence complete.  The 1960’s rang with smug, self-righteousness proclamations.  Many sociologists became involved as activists in the causes and movements of the period and chided those who did not.  We saw our colleagues masquerading their personal ideologies in the guise of scientific conclusions. This was a period of rebellion not only against archaic social institutions, but also against the traditions of science. It was a time for dogmatic promotion of counter cultures and “new sociologies”of this and that.  We are only now getting over the ecstasy of revisionism.  (Maybe some of us are not over it yet.)

The 60’s also brought new streams of research—but it was research with mixed blessings.  Coleman’s landmark study of test scores applied sophisticated methods and, and more important, it forced us to face up to the question of educational outcomes.  But we also learned from that study, I think, that test scores will not get us where we want to go.   And, while Coleman tried to include some measures of organization, the study was not founded on a theory of modern organizations. 

Well, there is obviously much more to the past than I can allude to here.  But, I’ve said enough to paint a picture of an immature specialty staggering back and forth between conflicting ideals, trying to find direction.  I believe that with maturity, the sociology of education will find its identity.  We can stop the shouting; stop proselytizing ideology; and stop bickering about revisionist frameworks.  We can stop thrashing around over psychological concepts.  We can discard our flimsy measures of organization and of education.   And, we can get on with the challenges ahead.

But will we grow up?  In my weak moments, I do think we are getting over the adolescent crisis and are about to embark on a more mature and ultimately more satisfying path.  Where will that path lead us?  We can find some clues in few basic conclusions that I am convinced we should by now have come to from our history.  So, I want to just briefly mention five conclusions concerning the discipline of sociology and four implications for research on education. 


Conclusions about Sociology

So, let’s start with conclusions about sociology.  First, I think that, by now, we have learned that individuals are not the province of sociology.  By individual, I am referring to the idea that we can study a person as a unique entity consisting of complex thoughts, motives, and emotions.  I want to make three points. 

ü            First, We can never study individuals as holistic entities.  Scientifically, the individual is only another abstraction, a set of selected attributes that can be observed and measured (such as age, gender.).  We can never study more than a few attributes.

ü            Secondly, thought processes, motives, and emotions are beyond the boundaries of sociology.

ü            Third, the attributes associated with individual abstractions are no more fundamental than any other set of abstract attributes.  Sociology is not derived from laws about individuals, and therefore cannot be reduced to individual actions.  Nor does it need to be reduced.  For example, we know that most people will seek to escape a burning building, just as most flies flee a fly swatter.  It is unnecessary to reconstruct the rational and irrational thought processes that humans and flies use to escape danger.  Another example: Disease spreads in crowed cities, notwithstanding motivations.

This point seems so clear.  And yet, prominent sociologists like Randall Collins continue to insist that only “real people” actually make events happen.  Alexander asked how it is possible to explain any social event without considering the individual.  Coleman advanced a convoluted model of social organization that begins with rational individuals making choices.  Other sociologists trotted out old “introspective” methods based on “empathy”– approaches that Mr. H. B. English long ago discounted as a “ghostly tradition”.

I am dwelling on this point—that sociology is not about individuals—partly because American sociologists have always been intrigued with social psychology.  And, more important, because psychology dominated the study of education until relatively recently.  So were have a lot of baggage.  Some researchers don’t seem to recognize any boundaries between psychology, social psychology and sociology.  So,

ü            We continue to insist on random samples of (you guessed it!) individuals.

ü            We embrace psychological concepts like “sense of control”, or “consciousness.”

ü            We conduct personal opinion surveys.

ü            And we are fond of topics like interpersonal interaction and socialization that are only remotely related to sociology.  

ü            But perhaps most distracting, we insist on “controlling for” characteristics of individuals before allowing organizational variables to enter our explanations – as though the attributes of people are somehow more fundamental than the organizations that make up what we call society.

            What do we actually learn by focusing on individuals, anyway.  Suppose, for example, we want to know about home schooling?  The natural inclination is identify the characteristics of the parents involved, their educational backgrounds, SES levels and racial/ethnic characteristics.  But what does such information tell us about home schooling?  We need to identify the organizations responsible for disseminating the program, establishing the curriculum and the procedures, and promoting the rationale.  And we want to know how parents are monitored, held accountable, and sanctioned.  We also need to find out what happens to various categories of students in identified types of programs—their career patterns, social participation and cultural activities.

But, you ask, why not include a variety of psychological, social psychological, anthropological and historical variables?  If you really believe Compte, that approach probably seems feasible.  But, if you acknowledge the highly specialized division of labor among the disciplines, you begin to suspect that the eclectic approach will lead only to more superficial research.  As sociologists, it is challenging enough tracing the organizational influences without dissipating energery on individual traits.  And, I do believe that with maturity, sociology of education will stand up to other disciplines with concepts and theories faithful to sociology.  Then, we will be in a position to contribute to an overall understanding of education.  But, if we go the eclectic route, we will not be able to devote the attention needed to develop a sound sociological knowledge base. We will continue to flounder around tracing personal variables without ever getting the organizational ones.

If individuals are not the province of sociology, what is?  This brings me to a second conclusion.  As Hawley and others have noted, organizations are the core subject matter of sociology.  Organizations are central to theories of the division of labor, social stratification, and social change and conflict.  In fact, I am not alone in defining sociology as the study of collective properties of organizations and relationships between them.  Sociology is about the interdependencies among organizations.

This thing we call education consists of vast networks of inter-connected organizations —schools, districts, state and federal agencies, systems of higher education, community groups, businesses, political parties, to mention a few.  We should sample these organizations and use them in explanations of important social outcomes. 

Easy to say.  But we face a major obstacle because prevalent methods often do not fit organizations.  For example, consider the statistical models that assume that sampling units are independent of one another.  Independent?  But clearly, schools and districts are tied into interdependent networks that include far-reaching government and private organizations. 

Similarly, we have to adjust our sampling techniques.  For example, we need to weight samples within organizations to reflect the power structure, rather than merely drawing random samples of the membership.  And, we also need to weight samples of organizations to reflect status hierarchies among organizations, and not merely draw random samples of schools.

            So, the point is that we face some hurdles because sociology is about organizations, whereas so many methods apply to individuals.  We need analytic methods compatible with the discipline.  But, as the discipline matures, I believe we will see a more concerted effort to use organizational concepts and theories to define problems and construct explanations.

The third conclusion we should have reached by now is that sociology cannot do everything.  We need to acknowledge our limits.  Compte told us that sociology is queen of the sciences, and following this logic, our specialty has a history of responding to all sorts of problems defined by practitioners and by philosophers.  Armed only with good intentions and blazing sociological arrogance, we have been quick to jump into the fray— to save burning cities, to improve low test-scores, to provide moral leadership for the nation. There is not a problem we have met that we couldn’t handle.  We have been a little slow learning to identify problems that can be addressed with our current state of knowledge and techniques. 

It seems to me that the best way to define problems is to first identify significant outcomes that can be explained with sociological concepts.    For example, we may want to know how a program affects the access that its graduates have to various types of schools and careers.  Or, we may want to know if a program influences participation in civic activities or political participation.

But what do we do?  Instead of starting with an outcome and tracing it sources, we fill journals with research on processes, such as classroom interactions and teaching styles apart from how these processes affect a given outcome.  We learn about many details associated with the “black boxes” without a clear understanding of how the processes relate to a significant outcome.  How many articles have you read that tell what outcome the author is interested in?

True, sometimes we do focus on one type of outcome: for example, standardized test scores. Which brings me to my fourth conclusion: research using standardized test scores has not been productive.  Our romance with test scores is a puzzling.  It is difficult to imagine an outcome that is less appropriate to sociology than standardized test scores. Tests are, after all, constructed to maximize variance among individuals within a classroom.  If an item does not discriminate among individuals, it is not included in the test.  By maximizing variance among individuals in this way, standardized tests are likely to minimize variance among schools, districts and other organizational units.  We need to use measures that maximize variance among the organizational units we are studying.

That brings me to a fifth conclusion, one that we have only begun to grasp.  It is this: Sociology is concerned with distributions, not central tendencies.   Blau told us, for example, that we should not be studying poverty per se, but poverty as part of the overall distribution of income and life styles. But truth is, we seldom focus on distributions.  We are preoccupied with central tendencies – again, partly because our analytic methods measure central tendencies.  To illustrate, I want to return to test scores despite my misgivings about them, because it will be easier to illustrate what I have in mind. 

Suppose a school has introduced a parent participation program, and a year later, mean test scores have not changed.  The implication is that it has had no effect.  But, of course this is not necessarily true.  Maybe students who were already at the high end of the distribution improved, but this improvement was masked by decline among students at the low end.  Or, the reverse.  The intervention helped low performers, but others suffered.

And, what if there is an increase in mean test scores?  That does not mean that everyone is better off.  The increase could occur in several different ways, each with very different policy implications.  For example: every student’s score rises equally; or only students at the top of the distribution improve; or students at the top improve substantially, while those near the bottom go down slightly.  Unless we know how each part of the distribution changes, we could reach misleading conclusions about the costs and benefits of this intervention.

To summarize to this point, I have identified five conclusions that I think we should have come to concerning the discipline of sociology.  Now, I want to illustrate some implications that these conclusions have for approaches we should be taking to research on a specific topic in education.


Implications for Research on Education

To simplify, I’ll concentrate on Charter Schools, which I have studied for several years. I will suggest four directions for research on charter schools, which will help to illustrate what I hope will typify the future of educational research. 

My first point, as you can guess, is simply this: we should use organizational theory to explain charter schools.  Charter schools are organizations.  And, as organizations, charter schools are negotiated entities.  Each district sponsor works out a distinctive balance between (1) the control it exercises and (2) the freedom it grants to charters.  To understand a charter school, or reach conclusions about its performance, we need to know about the balance of freedom and control.  Who controls it?  Which organized groups is it linked to?  How it is monitored and sanctioned?

Let’s focus for a moment on charter schools in California.   California law virtually dictates that every school must negotiate its own freedom.  On one hand, the law says that charter schools are supposed to operate independently.   But, on the other hand, districts (or other sponsoring agencies) are held responsible for approving, renewing and monitoring them.   How “independent” can a school actually be from a sponsor who is going to be held responsible for what that school does?  If sponsors are going to be held responsible, sponsors will impose their own requirements and interpretations. 

            Because each charter school must negotiate the balance between control and freedom, there are many kinds of charter schools.  Some schools are governed by districts, and do not function much differently than many other schools in the district.  But some do seem to be relatively independent of the district and of parents as well.  Principals and teachers (with little district supervision) govern some schools. And still others heavily influenced by businesses.  Some are controlled by small groups of parents, or elites.  Some are under the influence of groups with strong religious connections, or connections to ideological groups.  Also, some charters have converted from existing schools and carry that support and baggage; others are new and face the problems of many startup organizations. Obviously, each type of school has its own advantages and problems from the “git go”. 

The point is simply this: There is no simple answer to simplistic questions like: Are charter schools working?  You can’t generalize about “charter schools” as a category any more than you can generalize about public schools.  It is necessary to consider how each school is linked to other organizations.  In addition, since charter schools are supposed to be different from other district schools it is difficult to find comparison samples that make sense.  For example, a charter could adopt a program that deliberately takes longer to bring kids up to grade level in reading, or strive for outcomes alien to the district.

The second point I want to make is that as organizations, charter schools are vehicles of power.  We need to see charter schools as rivals for control.  One outcome we need to focus on, then, is how they do effect the distribution of power – in a community, state, or the nation as a whole?  And that is a complex issue.

For example: We hear that charters create options for parents, which give them more control over the education process.  However, parent control is often very illusive – because control can be compromised is several ways.  One way parent power can be compromised is that parents may not be well informed about the school.  A high percentage of parents in our case study of one school did not understand the Waldorf program the school was teaching.  They only bought into the general idea that charters are innovative.

Governing boards often create other obstacles to parent control.  Parents sometimes have no experience and often little time to devote to managing a new school.  And boards are political bodies, not always representative of the parent members. The principal and/or some teachers can step into the power vacuum and coopt the school. 

Another factor that changes the extent of parent control is selective recruitment.  But selective recruitment can operate very subtly.  It is not just a question of whether the school has a fair income and racial/ethnic composition – even though composition is often the focus of attention.  We conducted a study of schools that were using parent contracts.  That is, the school required parents to agree to specific commitments, such as donating time to the school, attending classes, or supervising homework. 

Our results suggested that schools often used contracts to screen out undesirable students—That is, students whose parents will not, or cannot, fulfill the requirements of contract.  These are parents who do not live up the ideal parent model as defined by the school.  It is true that the parents in that study who failed to meet the contractual requirements were disproportionately low-income racial and ethnic minorities.  But, the point is that selective recruit took place among the low-income group.

Think of it.  Here is a school serving low-income students.  But, through parent contracts, the school has been able to weed out some low-income students are the most difficult to reach.  It has a comparative advantage.  By comparison, a middle-income school that is not using contracts does not have the same advantage.  These conditions must be considered when comparing the performance of the two schools.

Another circumstance that I think will eventually erode charter school autonomy and parent control as well is simply this: some schools are skirting the law, or shall we say, treading the boundaries of impropriety.  For example, some schools are:

ü            Accused of teaching religion

ü            Operating in buildings that are only marginally safe

ü            Using practices that some critics say may be harmful

ü            Misappropriating public money

ü            Failing to keep parents informed or actually misleading them

The number of schools involved and whether they are actually doing bad things is beside the point.  The possibility is what matters.  We have to ask ourselves, what effect will these extreme cases have on charter school autonomy and parent control?  I think that, to the extent that schools operate independently of districts, hybrid bureaucracies will have to be created in order to monitor such excesses.  In other words, charter schools will be monitored.  If not by districts, then an agency will be created to do so.  Therefore, ironically, as an unintended consequence of freedom, in the future, there could be more public control over quasi-private sector schools because of the charter movement.

A third point I want to mention about research on charter schools is this: Standardized tests should be seen as strategy rather than outcome.  Charter schools can justify their freedom only to the extent they are somehow “accountable.”  Their sponsors turn to tests as a way to address the accountability question.  But the sociological question is not only whether students score better. The question is, how are tests used to legitimate and shape educational programs?  

We studied a charter school, for example, whose teachers felt compelled to include materials they considered incompatible with their program to assure that students could do well on tests.  And, again, the focus was on the central tendencies.  The district was examining whether the charter school had produced changes in the distributions, i.e., reduced or spread the range of achievement. 

We need in-depth studies of the complex organizational coalitions that are responsible for constructing tests, for monitoring them and for interpreting test results.  In other words, how do tests become institutionalized?  What groups are responsible? 

The fourth and final point I want to make about directions for research on charter schools is that  we should view them as part of larger social processes.  Charter schools – like vouchers and home schooling – are strategies.  They are strategies that powerful groups use to chip away at the monopolies that have traditionally controlled public education.  They are strategies toward privatization.  And, they are forcing realignments in traditional coalitions. 

            As such, Charter schools should be viewed as players in a social movement.  Clearly, we can make more effective use of the growing social movement literature that has been so illuminating in other areas, such as civil rights and violence.  We need to draw on that literature to explain what has been happening in education.  We should learn how programs like charter schools, vouchers, home schooling, as well as other movements, get started.  Which groups support them and how?  What kinds of networks form; how do they communicate? What social conditions facilitate the spread and the demise of programs like charter schools? It is more important for us to be studying questions like these than to focus so much of our effort on classrooms. 


Concluding Remarks

In conclusion, then, the sociology of education has the capacity to illuminate education as no other discipline can.  We should all strive for its unique contributions and avoid studies that duplicate what other disciplines can do.  Easy to say. But we need the settings and the incentives and well as the vision.  With that thought in mind, let me take just another minute before closing to comment on applied research.

We usually have had to make wrenching choices between academic and applied settings.  During adolescence, some of us ventured into government agencies, local districts, and applied research shops.  And, we usually found that it is difficult to met the demands of the discipline, which rewards contributions to knowledge, while working in more applied settings that seldom provide the time and freedom to stay abreast of the discipline. 

So what is the answer? I don’t know.  But a couple of things are worth mentioning.  First, I think we should work toward ways to institutionalize rotations of sociologists between academic and applied settings.  A second, and ultimately more promising tack, is to work toward the establishment of institutes based on teams and networks of sociologists, other social scientists, administrators, and practitioners.   A given institute could house teams of theorists and various types of applied researchers, and advisory task forces that include practitioners.  In this setting, sociologists can work with other social scientists to synthesize knowledge from different fields.  That seems more productive than the alternative—which is for each of us to try to comprehend and synthesize several complex specialties. The possibilities for institutes are more exciting when you consider the capacity of Internet connections to create virtual institutes—that is, collaboration among teams of colleagues physically located throughout the world.

Well, there are many other things I would like to say.  But my time—and your patience, I am sure—has run out.   Now, let’s hear from you.

 

 

Notes:

For whatever reasons, sociologists studying education have eagerly embraced individual-level measures, such as: test scores, vacuous measures of an individual’s “years of education,” studies of socialization, and mysterious “ethno-methodologies” designed to fathom mental states through self-analysis.  We are also fond of measuring the number of years of education an individual has completed, apart from the schools and programs involved. 

            Why the bias toward individual-level explanations?  I think the answer is a little bit of personal inclination and a lot of intimidation from other disciplines. On the personal side, in the name of “humanism”, some sociologists find it difficult to swallow the idea that human action comes from alien forces external to the individual. But, beyond that, education is a fiercely inter-disciplinary field that was for a long time dominated by psychology. But, there is no logical reason to focus on individuals and their motivations, since social causation occurs independently of motivations. Moreover, I don’t think that the preoccupation with blindly pursuing any variable that explains some variance will get us anywhere. 

 

            And, why don’t we focus more on organizational variables? One reason is that the tangible relationships among people within classrooms tend to be seductive. So, we like to study classroom interaction, for example. But what about re-organizing districts to improve teaching or promoting teaching relationships among teachers throughout a district or state?  Similiarly, we hear a lot about “parent involvement,” but very little about how to identify organizational boundaries, or how parents can effectively organize among themselves within a given school and among schools within a district or a state. 

            In our book, The Logic and Method of Sociology, Krishnan Namboodiri and I advocated an input-output approach to organizational networks.  This framework is premised on these three assumptions: 

(a) The focus of inquiry should be on a system’s output of goods and services. In service organizations like schools, which have vague goals, identifying the critical output is often a challenge.

(b) Outcomes are produced by sets of organizations linked into networks.  We must assume that the important factors affecting outcomes may lie in how the school is linked to the district and community.

(c) Outcomes are products of basic organizational processes, including competition, co-optation, coalition, bargaining, and conflict among organized power groups.  These processes impact outcomes as surely as teaching techniques and rulebooks.