Ib Ravn, Ph.D.Associate ProfessorLearning Lab DenmarkThe Danish University of EducationTuborgvej 164DK-2400 Copenhagen NVwww.lld.dk, -email- January 1, 2006, v. 2.0  (An earlier version of this paper was presented to the Danish Sociologists’ Congress,
Roskilde University, Denmark, August 18-20, 2005)

Abstract
Social or organizational theory usually represents an attempt to describe or explain phenomena or, at most, to criticize them. This paper argues that a theory can be transformative in the sense that in using it, researchers may actually help practitioners improve and transform their social institutions or organizations.  This idea is illustrated through the case of an R&D project carried out by the author’s group to help conference organizers develop meeting formats that create more learning among attendees than is accomplished by the conventional, lecture-based, one-way-communication format. A classically conceived theory of “the conference as a learning space” would describe and explain why conferences produce so little learning as they often do, but is this good enough? Might not researchers play a more active role in helping practitioners transform their domain? We devised a theory of learning at conferences in terms of human co-flourishing (Aristotle) and articulated four design principles that would help conference planners design better conferences. We then fashioned 15-odd interpersonal activities intended to improve learning for the conference attendees, and we implemented each of them in some of the 32 conferences included in the project. Deriving from the design principles and the theory of human co-flourishing, each activity thus constitutes an hypothesis that was tested in practice; a hypothesis that when attendees do perform activity X, their learning will be improved. We collected data from four data sources to ascertain whether learning had indeed improved for each of the activities tested. We are currently in the process of writing up our results, some of which are about individual activities, others modify our original four design principles, and others question the underlying theory.  The project illustrates the use of social and organizational theory as transformative. Such a theory describes not a current mess, but a future state that does not have the problems found in the present. Specific hypotheses suggest how that future is to be approximated, and trying them out in practice will inform our knowledge of how to change the domain.  The paper concludes with a sketch of a seven-step research cycle that compares transformative research with classical hypothetical-deductive methodology. They are structurally very similar, but one is concerned with a messy present that the researcher can do little to help change, the other with a desirable future that the very research effort is intended to help bring about.
 
Introduction
In 2003, Learning Lab Denmark was invited to give expert input to a group of executives from leading Danish hotels and congress centers. The group was assembled by the industry organizations VisitDenmark and Wonderful Copenhagen, and the executives were concerned that the conferences and large meetings that make up about half their business were run in much the same way as thirty years ago. No meeting innovations had seen the light of day since the flipchart and the overhead projector. The executives knew that their customers were bored by the traditional string of endless PowerPoint presentations and they wanted to explore “meeting concepts of the future”. Based on our knowledge of learning theory and educational practice, did we have any ideas on how to design and run a “learning meeting,” as they called it?  Our response, in the affirmative, led to a research-and-development effort (Ravn, 2005), ongoing at this time, during which we have had occasion to try out a type of social or organizational theory that we call transformative. A theory of this kind describes not a present (and problematic) state of the world, as many a social or organizational theory does, but a desirable future that has none of the problems found in the present. The theory is transformative in the sense that it is intended to help the stakeholders of a system or a domain transform their practice and institutions in such a way that the problems they experience at present do not arise. In the conference case, given the widespread complaints about boring conferences, we were not particularly interested in a theory of conference planning or attendance that lets us measure, describe and explain what average conference organizers and attendees do currently. We were more interested in a theory of conferences that would help organizers put on terrific conferences that provide superior inspiration and professional development to everyone present.  This conception of what a social-scientific theory may be and how it may guide research and practice grows out of earlier considerations of mine (Ravn, 1986, 1991, Baburoglu and Ravn 1992). My chief inspiration is the work of the American management thinker and co-founder of operations research, Russell L. Ackoff, in whose gradual career shift from social research methodology (Ackoff, 1953) to corporate planning (Ackoff, 1981) I found commonalities (Ravn, 1986) pointing to a very different view of what a social and organizational theory could be: The coherent identification of a desirable future and the means to bring it about through systematic implementation and experimentation. Of course, broadly similar arguments that social research should be so conceived that researchers contribute directly to organizational and institutional development have been advanced by action researchers (Lewin, 1947; Susman and Evered, 1978; Reason and Bradbury, 2001), students of professional learning (Schon, 1983) and organizational learning (Argyris, 1999), social constructionists (Gergen, 1982) and others too numerous to mention. Lately, Gibbons and co-workers (Nowothy et al., 2001) have discussed the increasing interweaving of research and action in the knowledge society, identifying a mode-2 type of research that sets knowledge in the context of application. Medically inspired intervention research is very explicit about “administering treatments” to groups and organizations of people (Rothman and Thomas, 1994), while quasi-experimental research designs work around the taxing requirements of randomization in an attempt to determine which interventions work (Cook and Campbell, 1979).  To these varied schools and authors that struggle to make social research useful to society I wish to add a bit of pragmatic philosophy and some humanistic science. This will enable us to specify much more directly what we may require of a social-scientific theory for it to be relevant to social and organizational transformation. Rather than specify at length where I part company with the writers just mentioned, led me take the practical route and introduce transformative theory through the case of the learning conference.   I will show how an organizational domain (in casu, conferences and large meetings) that is mired in problems (conferences tend to produce very little real knowledge sharing or learning) may benefit from a systematic and experimental research-and-development effort. Central to this effort is our transformative theory and an associated set of four design principles for learning conferences. From these principles are derived hypotheses for better practice, and they are tested in concrete conferences: do attendees actually learn more and get more out of their attendance when we change the conferences as indicated by the design principles and the transformative theory? I then sum up the concept of transformative theory and show that it can be used productively when informed by scientific, experimental methods. We can experiment with different kinds of conference designs and compare attendee eval­uations, thus improving on theory as well as practice in one integrated process.  Finally, I explore what it means for a social theory to be thus transformative and prescriptive, arguing that since the human-social world is subject to ongoing construction anyway, researchers might as well admit up front that they partake in this construction through their research efforts and, indeed, use transformative theories to try to help social systems and domains perform more to the satisfaction of their stakeholders.
Helping clarify what better conferences are
In response to the request made by the hotel executives, our group at Learning Lab Denmark set to work. We determined from informal interviews and our own experience what was wrong with the typical one-day conference for business people and professionals (cf. Ravn, 2005). Apart from the overwhelming one-way communication, conferences generally do very little to involve people or help them interact meaningfully. A meager five minutes for Q’s and A’s after a presentation is common. The members of the audience who step up to ask the questions are often so starved for attention that they will ask long-winded, excessively idiosyncratic or “See how clever I am”-questions.  If there is a panel of experts in the afternoon, all of them will generally feel called upon to respond to each and every question from floor. If the attempt has been made to activate attendees in the form of group-work, the groups are usually left to their own devices. This is an invitation to the most active speakers to hijack the conversation while the rest of members tune out and often end up wasting their time.  As the next step, we searched the literature for alternatives to the PowerPoint-ridden conference, mostly in vain (Open Space Technology (Owen, 1997) being a notable exception). We asked experienced facilitators and process consultants what they did with a hundred people gathered for a meeting. We designed a number of knowledge-sharing processes ourselves and tried them out on clients. We presented ideas and principles for “learning conferences” to the executives, eliciting their feedback. And we wrote up a small manual for hotel personnel intent on helping clients put on conferences more conducive to learning (Ravn and Tange, 2004).  Having been thus inspired, we saw an opportunity for pushing ahead with a more carefully designed research and development effort that would deepen and consolidate our initial explorations and ideas. As we discovered, meetings and conferences are a multibillion dollar industry worldwide (Rogers, 2003), and the dozen Danish hotels in Rezidor SAS Hospitality pull about half of their income from meetings and the night stays they generate (Sorang, 2004). The Danish Ministry of Economics and Business Affairs had their eyes set on the meeting industry, hoping to help hotels and convention centers attract more international congresses to Denmark. The concept of a “learning conference” is very much in line with Danish traditions for progressive and egalitarian schooling, lifelong learning, informal power structures, participation, worker-management collaboration and so on. Could the Learning Conference even become a Danish brand? At Learning Lab Denmark, we picked up the challenge and designed a project that would explore the meaning of the learning conference through an 18-month process of experimentation and reflection. We recruited partners in the meetings industry: Rezidor SAS Hospitality, Hotel Legoland, Nyborg Strand, Odense Congress Centre and one large customer, Danske Bank. We obtained partial funding from the Ministry and the rest from the partner companies.  A team was put together: My colleague Nina Tange would facilitate, as would  Malene Rix, an independent process consultant we contracted with. They would help the hotels and conference planners change the design and execution of already planned conferences so as to make them more “learning”. A researcher, Steen Elsborg, would evaluate their efforts through observation, stakeholder interviews and surveys of attendee satisfaction. The present author would provide coffee and donuts and tie the efforts together qua project leader. The researcher would thus operate in some separation from the change agents and hence not be forced to evaluate his own intervention efforts (a very difficult position that action researchers and research clinicians like psychiatrists and psychoanalysts often find themselves in). Yet, we would work as a team and together advance all parts of the research-and-development effort. At the moment of writing (mid-2005), we are almost midway in the string of 32 conferences lined up for the project. For each conference, our facilitators offer help in making the event yield more learning for attendees. Facilitators help organizers redesign the original plans for the conference, introducing better processes for attendee interaction and knowledge sharing, fewer and shorter presentations, safe small-group conversations for attendees to speak their minds, short quiet periods for individual reflection after presentations, richer interaction between attendees and lecturers during and after presentations, instructions for group leaders to facilitate group discussions, breaks with gently guided processes for interaction, special buffet lunches that induce attendees to circulate and talk to strangers, dinners where guests can legitimately escape the bore in the next seat and meet other people, etc., etc. Future publications will describe these processes and what we learnt about their effects.
Why a descriptive-explanatory theory would fall short
Now, backtracking a little, we entered this domain in much the same way as the researchers or consultants among our readers probably have done occasionally: As experts on broad principles and neighboring domains (learning, knowledge sharing, group facilitation), yet having little intimate knowledge of the specific topic of conferences or the meeting industry. Responding to an expressed need for change, we looked around for ideas and practices that would help our clients and partners improve their practice, that is, put on better conferences. Yet, we are researchers, not consultants, so we obviously wanted our efforts to produce knowledge, not just change or business development.  At the outset, the obvious social-scientific research avenue was open to us: Go in and describe the domain, measure its variables and try to create a theory that will model, explain or maybe even predict behavior. This would probably not have been too difficult, as conference planners on the whole appear to be a pretty predictable lot. They seem to feel that since they are troubling people to go to the conference, they are obliged to put on the best show possible: Lots of intense presentations, so none of the attendees’ precious time is wasted on chit-chat. Add some entertainment – celebrity lecturers, events, team-building – to compensate for the boredom that the organizers sort of expect the presentations to produce, but which they believe is unavoidable.  Underlying current practices in the meeting industry is, of course, some notion of what it is to communicate and share knowledge at a conference. Scratch a conference planner’s brain and the transfer model of teaching will pop up: Knowledge is held in the brains of experts who can transfer it to other people’s brains by saying it out loud from a podium, supported by visuals like PowerPoint slides. The assumptions about human behavior associated with this model will go a long way toward predicting the behavior of conference organizers.   In constructing such an explanatory theory of conferences, we could throw in some observations about the behavior of conference attendees. On the whole, they are well-behaved and will sit in their chairs and wait to be lectured at or entertained. They will take in what they can and then stop. Studies of human attentional capacity may help predict what the threshold point is, but a more informal method is to observe how many people actually look at the presenter or how many slip out gradually during the afternoon presentations.  If we were to include the attendees in our description, we could elicit their reasons for going. A critical researcher would have a field day identifying rationalizations of a magnitude proportional to the costs involved in attending: It takes guts to admit that going was a mistake. On the other hand, many conferences do have attendees that thoroughly enjoy being there, and there is a relevant challenge for researchers to discover what factors in this learning space are indeed associated with attendee learning and other desired outcomes.  In other words, it is entirely possible to create a classical, descriptive-explanatory-predictive theory of the conference domain; a coherent set of propositions that yields understanding and is testable against empirical reality. The theory might be purely descriptive, or it might be terribly critical, exposing the power interests of conference hosts who may not want the rank-and-file attendees to speak up at all, or it might reveal the socially constructed character of conferences, intimating that they could indeed be very different (but not telling precisely how); or it could be feminist, discourse analytic, phenomenological, take your pick. However, none of these versions of theory appealed to us. Common to them all is a preoccupation with the present and a reluctance to entertain the future, let alone be very specific about it. Our hotel partners and we had a reasonably good idea of there things stand in the conference world. A research project that described the present rut and concluded with the conventional list of seven two-sentence ideas for possible alternative conference formats would have suited neither our aspirations nor those of our partners or funders. We wanted a more proactive theory, a kind of systematic understanding of the domain that would help us and our partners create better conferences, pure and simple. Recommendations for action would have to flow directly and explicitly from this theory, and not be the anecdotal afterthoughts or isolated bright ideas that researchers often append to their reports.
A theory of human co-flourishing for the learning conference
In preparation for project “The Learning Meeting” we reworked the three design principles for learning conferences that we presented to the hotel executives originally. We added a fourth principle and massaged them all into a humanistic view of knowledge sharing that was emerging from our office conversations, as well as a pragmatic view of knowledge developed in another context, for intellectual capital statements (Ravn, 2004). As the reader will agree, theories do not spring to life fully-formed, like Athena from Zeus’ forehead.  In any case, the conception we now call the theory of human co-flourishing is in harmony with broad humanistic and pragmatic ideas about knowledge and knowledge sharing (or learning), from which our four design principles may be understood as being derived (though, chronologically, they were not, but that’s the way “logical” derivations go, isn’t it?). The transfer model of teaching presumes the individual to be a Lockean blank slate, a Skinnerian black box or an empty vessel to be filled with the teacher’s knowledge. Of course, modern psychology and educational research have documented that people are far from blank slates. Humans are born with a set of predispositions to walk, to talk, to think, to be active and contribute, etc. Long unfashionable, theories of human nature are resurfacing (Pinker, 2002; Seligman and Csikszentmihalyi, 2000) They combine with the Aristotelian idea that is the purpose of human existence lies in human flourishing (Paul, Miller and Paul, 1999), that is, the unfolding and realization of human potentials and talents. Human nature can be conceived of as innate needs, motivators and potentials for action. People are motivated to do for themselves and for others what they do best, they wish to grow, develop and flourish.  On this view, to learn is to be better able to unfold human potentials. Learning is not narrowly about acquiring a stock of knowledge. Learning is learning how to act so as to increase human flourishing. Likewise, we dispense with the traditional philosophical position that knowledge is representations of an external reality stored in the mind or in computers. Knowledge resides in human activity and shows itself in optimal human action. A knowledgeable accountant is not one who has memorized thousands of facts or rules about accounting, but one who in her professional practice can help her clients do their accounts in a superior manner (Schon, 1983; Brown and Duguid, 2000). Knowledge is the structures in individual or collective consciousness – such as categories, distinctions, concepts, images, etc. – that help people act better (James, 1907; Dewey, 1938, Argyris, Putnam and Mclain Smith, 1985) and thus fulfill their potentials as human beings (Rogers, 1961; Ravn 2004). Re-enter the conference. If this is human nature, how may we then conceive of a conference? Well, we could see it as a forum where smart people present their knowledge and other smart people wish to be inspired. Let us drop the idea of the conference as a place for one-way communication, inefficient as it is, because going against most of what we know about natural learning in humans. Rather, let us see the conference as a forum for shared learning, mutual inspiration and knowledge sharing – a collective learning space (Kolb and Kolb, 2005). Let us suppose that people go there to be inspired and learn new stuff that will help them do what they want to do with their skills, talents and potentials. People are active by nature and will want to be active during conferences, as well. Unlike empty vessels, conference-goers are brimming with knowledge, preconceptions, inclinations, ideas, projects and things they want to accomplish. We can safely assume that most conference attendees have just barely managed to take a day off from their busy schedules to go find inspiration about a topic they care deeply about (at least if they go voluntarily).  Now, as to the conference program, there must indeed be presentations, speakers who address an audience, or we would call it a meeting, a network gathering, a reception or a party. But since presentations are a major source of boredom and failed learning at present-day conferences, they must be better. From experience, let us suggest that they must be brief and present original and provocative ideas, and there should be far fewer of them. Let us formulate that as a design principle, number one: 1.    Presentations must be few, concise and provocative. Next, audience passivity is out. Plenty of research has shown that passive listening leads to very little knowledge acquisition, attitude change or behavioral change. So the audience must be involved in an active interpretation and acquisition of what is being presented. Since we judge everything we hear by what we know already, the conference must utilize this fact to help people digest the input presented. Also, since people want to do things with what they hear (use it or lose it, as the saying goes), the audience must be led to reflect on and talk about how the input may help them do what they want to do already. The conference must provide processes and questions that help people reflect on and talk about this. Hence design principle no. 2: 2.    The conference must provide processes for attendees to engage in active interpretation of the input. Since people go to conferences on topics that concern them greatly already, and since learning is most efficient where you are highly motivated and in domains you are just entering and are ready to gain control over (Vygotsky, 1978), it makes sense to give attendees a space to talk about their relevant concerns and projects at the conference. Join them in pairs and triplets and ask them to talk about their burning issues, just for a couple of brief periods during the conference. They will feel gratified for the attention they get from the others, and they will listen and learn from other folks who are excited about the projects that fill their professional lives. Thus design principle no. 3:          3.    The conference must provide processes for attendees to talk about their own relevant projects.

Finally, there are other interesting people at the conference besides me and the presenters. In fact, in the knowledge society, we could argue that the real resource at conferences are the other attendees, many of whom have as much to offer as the experts on the podium. Already, many people acknowledge that meeting these other people is a major reason for going. Why not work that into the conference design much more explicitly than the lazy routine of just providing breaks and free time? There is tremendous energy, excitement and learning to be had from meeting people out of the blue who are committed to just the same issues as oneself – or who, indeed, have slightly different take on them, enough to make you change directions and strike up a partnership with them. Hence, the last design principle: 4.    The conference must provide processes for networking and knowledge sharing amongst all attendees.  So, for conferences to be fora for learning and human co-flourishing, we must design forms of communication and interaction that will allow people to listen to presenters and each other and to share their thoughts with many others, all for the purpose of facilitating the kind of mutual inspiration and motivation that will help everyone work for their goals, fulfill their potentials and flourish together as human beings. Of course, I wouldn’t claim that the design principles presented are the only ones that can be derived from the theory of human co-flourishing, nor that they couldn’t be derived from some theory of a different slant. But I feel confident that the reader didn’t jump in her seat, nor uttered “Nonsense!” under her breath during her reading. It all hangs together reasonably well for a first piece of research into conferences.
Testing hypotheses = transforming social practices
Let us ask in what sense these ideas and principles constitute a theory, a social-scientific theory. Well, informally, they are a theory of how things could be much better than they are at present, as opposed to a theory of what things are right now. As action researchers Peter Reason and William Torbert (2001) put it, “…good action theory will offer a normative vision of a better state”. Such a theory is testable through attempts to bring about the conditions it stipulates. The theory may be deemed good if attempts to implement its various elements in actual practice meet with success, as judged by the stakeholders involved, using appropriate standards. And the theory fails to the extent that its fails to produce these desirable outcomes. A first step to making the theory amenable to implementation is to spell it out, that is, isolate elements of it that predict very specific outcomes. For example, the part of the theory concerned with the attendees’ need to engage in active interpretation of the input presented, that is, design principle no. 2, may be fleshed out as the expectation that if we let attendees discuss the presentation with the person sitting next to them for five minutes midway through the presentation (a so-called humming dyad), their outcome of the presentation will be greater; they will have learned more. This expectation may be tested in an experiment: when the conference host breaks a presentation and asks attendees to do just this, will they afterwards judge their learning outcome from the presentation favorably? Obviously, expectations thus derived from a conceptually tight framework and tested in actual experiments may, without stretching the term, be called hypotheses. Not that the testing of the hypotheses through implementation and evaluation is at all easy, but that is a challenge common to all research. The results produced during the testing of the hypotheses will feed back into the design principles and the theory, thus helping us refine both theory and design principles.  For example, in the research project mentioned, the repeated use of humming dyads have led us to refine its specification: If the couple is free to discuss anything, they will tend to share critiques of the presentation, and this seems to produce less learning and poorer outcomes than we expected, as judged by the attendees themselves. This has led us to propose that the humming dyad must be introduced by the conference host asking a constructive question, like “What inspired you the most in the presentation so far, and why?” Our refined hypothesis, yet to be tested, is that this will take their minds off facile, academic critiques and focus their conversation on the presentation’s learning potential for them personally, thus improving their outcomes. Such are the very useful reflections occasioned by attempts to implement the various elements of a transformative theory, that is, to test hypotheses derived from it. The theory is not descriptive or explanatory in the classical sense, in that it does not attempt to explain conferences as they are now. Rather, it is transformative, in the sense that researchers and practitioners will use it to try and transform current practices and institutions into something better – as forecast by those fashioning the theory and as judged by the stakeholders trying it out.
Values and human being
Now, let’s face the issue of how a theory can say what’s better and what isn’t. It has hardly escaped the reader’s attention that transformative theory as depicted here is rife with values. When I described the theory of the learning conference, I went from an account of human nature and general human capacities and straight to ideas about what a conference that would maximize learning would be like. My guide was, obviously, that conferences should help bring out the relevant aspects of human potentials, that is, our potentials to learn. For conferences to help us do this, they must have the features identified in the four design principles, I argued.  So, we have a (purportedly scientific) theory that says that conferences that conform to our design principles are good, and conferences that don’t are bad. That doesn’t sound very scientific. But let’s go closer. A transformative theory works on the assumption that bringing out human potentials is good. This is the humanistic basis of transformative theory. While this may sound fine, a critical reader may well ask how we distinguish our “good” potentials from our “bad” potentials. Indeed, we would not wish for at conference to realize our potentials for being bored and wasting our time. So how to distinguish? An answer will take us beyond the bounds of this paper. Elsewhere, however, I have argued at length for a humanistic basis for conceptions of the good life, and I have dismissed the frequent (and fallacious) counterargument known as the naturalistic fallacy (Ravn, 1989).  For here, suffice it to say that researchers in many other sciences have little difficulty seeing themselves as working a common human good. Take pedagogy, human development, pediatrics, educational studies, nutrition science, biomedicine, nursing, family medicine, cognitive science, psychiatry, clinical psychology and organizational psychology. They all proceed from the assumption that human nature can be studied, and what we learn must help us act in such a way that we bring out the best in human nature.  Social scientists will argue that society is not given by human nature. Granted, no direct derivation can be made; blueprints for optimal social institutions cannot be read out of the human genome. But does that amount to saying that nothing in human nature has relevance for the way we live and work in society? Clearly not. Sure enough, social scientists have been extremely wary of arguments that certain human qualities or condition are natural and therefore should be favored in society. Thus, fifty years ago, scholars in the field of International Relations would laugh at the idea that as scholars, they could condemn genocide and similar internal affairs of other states. Political scientists couldn’t argue, as scientists, that democracy or human rights are a good thing. Legal scholars emulating the objectivity of positivist science could not declare the atrocities occurring in Germany in 1942 to be unjust (Høilund, 1992: 45). But all that is changing. Human rights and human development are increasingly seen as being for all people, and broad humanistic values are creeping into the social sciences. Human nature, human needs and human potentials must be acknowledged and their expression facilitated, and social scientists today are much more likely to accept their part in this than fifty years ago.  Of course, no one can declare ex cathedra that practice X is better than practice Y. This is the province of ideology, systems of fixed ideas. Contrast this with the scientific method, which, while being likewise concerned with systems of ideas, focuses on their continuous improvement through testing, experimentation, evaluation, reflection and learning. Thus, the claim that practice X is better than practice Y calls for an experiment. Let’s try them both out and evaluate them afterwards.  This is commonplace in matters of social and other policy, where social experiments set up by program administrators are tested every day. When politicians or bureaucrats decide to institute some labor training program, they often hire a researcher to do the evaluation, using whatever criteria the domain has to offer – all of which boil down to one version or another of “good for people”: user satisfaction, contribution to well-being, quality-of-life indices, subjective health, etc., including economic indicators like cost efficiency, which amount to: “Can we get more user satisfaction out of spending the money differently?”. So having researchers estimate whether stakeholders judge a program or an action to be good or bad is nothing unusual. However, placing the values directly in the theory is unusual, indeed. It is almost anathema, a contradiction in terms, logically unthinkable. But this is mainly because the sciences have concerned themselves with actualities, not potentials, including human potentials. Actualities are factual and devoid of value, they just are; whereas potentials come with values. It is difficult to say “potential” without implying that it has to be realized. It is time to include potentials in the social sciences, that is, the potentials that social systems and domains hold for becoming much more conducive to human flourishing. As Chris Argyris states: “A complete description of reality requires not only a description of the universe as it is but a description of its potential for significantly reformulating itself (its potential being part of what it is)” (Argyris, 1982: 469). Potentials fell out of favor when early science gradually replaced medieval Aristotelianism. Aristotle was big on potentials; fire would seek upwards because its natural home was there, and it fulfilled its potential by moving up. Such fantasies made a poor name for potentials in the sciences. Just as fire was later recognized as combustion, a simple chemical process with no preferred directions or intrinsic values, so social phenomena captured by the “physique sociale” of the early 19th century, or “sociologie”, as Auguste Comte in 1839 chose to call it instead, were to be described impassionately, for what they are here and now, without regard for values or preferred directions or improvements.  In the mechanistic science reigning between 1850 and 1950 – just the period when the social sciences came of age – the human being was seen as a machine, a Skinner box, behaving like a sequence of stimuli and responses. There was no room for a human nature with intimations of greatness, potentials waiting to be discovered and made to blossom, natural needs crying for expression and satisfaction. This came later, with the humanistic psychology of Rogers and Maslow, progressive education and reform pedagogy, the social reforms in the welfare state and the alternative lifestyles of the burgeoning consumer society. Having made its impact on everything from infant care to the American presidency (notice the change in outlook from old-school Reagan and Bush Sr. to baby boomers Clinton and Gore), the humanistic turn in matters social and human has yet to make its impact felt in the very particular domain of interest here, social research methodology. On the whole, the social research methods available keep social scientists in their academically respectable roles as cautious and critical analyzers of society. Pick up any handbook of social research methods: They are all about describing and looking critically at social life, not about helping change it for the better the way scientists could if they really would – the systematic, theoretically sound, experimental, evidence-based way. There is a glaring dearth of methods that let researchers use their full intellectual powers and act as designers and facilitators of real-world experiments with practices and institutions that promise to bring improvements, however small or uncertain, to the quality of people’s lives in society. Let’s refocus on how this may be done in practical terms.
A transformative research cycle
A transformative theory is but one part of a research process. Since we have used familiar terms like theory, hypotheses, testing and modification of theory, let us line these elements up in the kind of diagram of the research cycle that has fallen out of fashion in the social sciences over the past couple of decades (because who wishes to take a stand for any scientific method in these post-structural times?). I wish to compare the present approach with research of the classical hypothetical-deductive kind.  1.    Research starts with a problem, traditionally one of understanding. The researcher doesn’t understand why the domain behaves the way it does, or does not accept the prevailing understanding. From the transformative point of view, however, the problem is one of practice. Things in the domain don’t work as well as they should; there are stakeholders who wish to act differently and see improvement. In the pragmatic tradition, this starting point for research is commonplace (Dewey, James).

In our case, the domain is conferences. The practical problem: They seem to be ineffective vehicles for learning. What can we do? Also, the hotel executives were concerned abut their business: how to keep our customers coming for conferences if their attendees find them so boring? 2.    A theory is fashioned by the researcher. Unlike the descriptive, explanatory or predictive powers that a classical researcher may wish for her theory, transformative potency is what we shoot for. When applied and tested, will the theory likely lead to transformation and improvement in relevant social practices and institutions? To accomplish this, the theory must depict the potentials and possibilities inherent in the domain, pointing out what could become reality if things worked out just right and all relevant actors conspired to do their best. Still, it is a social-scientific theory, not an ideological tract for unthinking revolution, so it must adhere to the conventional standards of rigor, coherence, parsimony and agreement with other research-based knowledge about the domain.

Our transformative theory is the idea that the conference is a forum for human co-flourishing, where people go to find inspiration to bring out their best potentials. The four design principles spell out what this implies for presentations and attendee involvement and interaction.

3.    Derive hypotheses is the next step. That is, make the theory concrete and applicable in some local domain. We don’t want to test whether explanations or predictions are true or false; we want to use our transformative theory as a guide to the action that our stakeholders wish to take. The hypotheses are these more specific guides: When you do X in situation Y, we expect outcome Z to be produced. Social experimentation being something of an art, these hypotheses are generally far from well-formed and would hardly meet the requirements of positivist research methodology. But less than that will do. They just need to be reasonably clear and distinct expectations about the results of action, put in words and written down before the actions are taken – so as to prevent post hoc rationalizations and thus facilitate learning.

Before each of the 32 conferences our facilitators meet with the relevant conference organizers, listen to their needs and negotiate techniques and designs that both parties hope will deliver better outcomes. Each design element or meeting techniques is chosen for a reason, which may be expressed as an expectation about the good outcome it will produce. These expectations are hypotheses, and they are noted in a log before the conference.

4.    Act as specified by hypotheses. In the conventional track, there is no action. A researcher forms a hypothesis (the previous step) and collects data to see if she was right (the next step). In the transformative track, however, we are not concerned with passively knowing more about the present, we would like to know – in very active and pragmatic terms – how to make the future better than the present. So we act as prescribed by the hypotheses. We do X in situation Y, hoping to obtain outcome Z.

In every one of the 32 conferences, we execute a number of meeting processes, which, as expressed in the hypotheses, are intended to improve learning and other desired outcomes.

5.    Testing and evaluation. Conventional data collection and analysis will
determine the fit between hypothesis and reality. In transformative terms, we ask if the processes specified by our facilitators produce the desired outcomes, individually or collectively? Obviously, effects of the different actions and interventions are difficult to tease out, and the usual problems of (especially internal) validity are present here in abundance.

Evaluation is the job for Steen Elsborg, researcher. Surveys ask attendees to evaluate the processes that went into the conference; interviews allow him to probe deeper; participant observation of the 32 conferences reveals things no one else notices, and continuous conversation, interpretation and reflection with the rest of the team and with partners during and after each conference yield further insights. Some of these insights are used by the facilitators to change their further practice in the projects, much in the way that action research and other collaborative and reflective kinds of research allow continuous learning and improvement during the project.

6.    Generalization . Traditionally, researchers decide whether the theory needs to be modified in light of the hypothesis testing, so as to increase its external validity (generalizability). Using a transformative approach, our ambitions are the same: we would like to know whether the actions specified in our hypotheses would lead to the same results in other situations. This scientific ambition mixes with our pragmatic intention to improve behavior in the domain at hand. These two concerns are not easy to reconcile, especially because to be effective in a concrete situation, one has to make innumerable adaptations that do not necessarily translate to other situations. However, the fact of this conflict is trivial; it is known from every action we take when we try to learn afterwards what can be done better next time. Likewise, using a transformative theory for research is a repeated process of action, reflecting and learning.
  Common
feature Hypothetical-deductive method  Transformative method(exemplified by “The Learning
Conference”) Difference between the two methods 
1. Topic Identify interesting topic Identify interesting topic (Conferences and their poor learning) Does the interest relate to a problem of understanding or a practical problem?
2. Theory Find or create
descriptive or
explanatory theory Create transformative theory(People wish to grow and learn, so conferences should be learning and conform to the four design principles) How things are right now vs. how we would like them to be
3. Derive hypotheses Specify what we expect to find when we look at the world through our theory Derive processes that would help improve performance in the domain(Design new meeting techniques and specify what we expect them to accomplish) Do hypotheses seek to describe the domain or help change it?
4. Act according to the hypothesis (not applicable) Try to improve things in practice(Facilitator tries out the new meeting techniques) No real-world action required vs. transformative action
5. Testing, evaluation  Collect and analyze data: Is the hypothesis supported or falsified? Did things in fact improve, as judged by stakeholders and researcher?(The researcher identifies which processes had positive effects on learning) Resistance to falsification vs. practical usefulness
6. Genera­lization Modify theory to increase its generalizability, veracity, truth (go back to step 2)  If the action did not produce the desired effect, why not? (Go back to step 2). If it did, would it in other situations, too?  (What meeting processes will work at other conferences?) A good theory is a true representation of external reality vs. a good theory helps us act effectively in many situations
Figure 1. Comparing the hypothetical-deductive method with the transformative research process.   Of course, in practice, these six steps overlap, feed back, repeat, etc., etc., as in all research. The main thrust should be clear, however: our concern is with identifying potentials for action and to act accordingly, learning as much as possible about our prospects for further action again. Transformative research exploits many of the elements of classical research, but modifies them to suit the underlying purpose of identifying realistic trajectories for human and social development and start a systematic, reflective, experimental process with major action-theory interaction, all designed to help improve the domain and the quality of its stakeholders’ lives.
Research vs. change agency
It may be argued that the process described is nothing more than regular, sound advice for practical action. Isn’t this what any change agent, consultant or manager does intends to do? Every stakeholder can dream up a better state for his system, try to act to bring it about and do some evaluation along the way. Isn’t this just a recipe for offhand social experimentation of the kind that grassroots activists like to advocate? Why call it research? In response, consider the following features that distinguish the transformative research process from ordinary, well-informed action:  ·         General theory
There is a theory that purports to cover the domain in general. A change agent’s primary desire is to have his own organization work and cares more for particulars than generalizations. ·         Coherent ontology
The theory is grounded in an ontology, a wider understanding or coherent view of human nature, social institutions, democracy, justice etc., that must be in agreement with (at least some) scholarly thinking. We rule out artistic-creative, ideological, religious or private visions that are completely out of line with what is known about human and social possibilities and potentials. ·         The theory yields testable hypotheses
A transformative theory identifies potentials for human and social development. Precisely how these potentials may be realized is what we express in a hypothesis. A hypothesis is the conjecture that “If we do X in situation Y, effect Z will occur.” Such a hypothesis is testable: Do it, and see if Z happens.. Now, comparing with managers and activists, it is true that they often operate from a mission or a vision of how things could be, akin to our transformative theory. Yet, the prescriptions for action derived from a corporate mission are rarely couched as hypotheses to be tested in experiments.  ·         Systematic evaluation of hypotheses
The actions taken are evaluated systematically by researchers and the domain’s stakeholders. Did they produce the expected and desired effects? Managers and change agents are usually too busy to spend much time systematically evaluating the changes they introduce. To the extent they do, we see them as scientifically minded, because evaluation through data collection and analysis is integral to any scientific method. ·         Evaluations feed back to improve theory’s generalizability
 The evaluations produced by researchers and stakeholders are fed back to improve and modify the theory, thus making it more precise in its applicability. If hypothesis/action X leads to favorable evaluations by the stakeholders of system Y, we are encouraged to believe that this action will work in other, somewhat similar systems. But if action X produced negative evaluations, we may wonder if we need to stop doing X altogether or maybe do it in a very different system. Evaluations help us explore how widely the theory holds. Managers and change agents, on the other hand, will wish to use evaluations to learn about their particular system, caring less about “Where else would it work?” A concern with generalizability is a hallmark of scientific research.·         Theory-action consistency
Through the medium of the testable hypothesis, actions are kept aligned with theory. While consultants and managers will happily change strategy the second it is called for, a transformative researcher has an obligation to theory or general understanding and wants to preserve the integrity of theory or at least modify it with eyes wide open. Theory must lead to effective action, but action cannot proceed unthinkingly.  In sum, transformative research is not just well-informed action, as it stresses theoretical comprehensiveness and generalizability, theory-action consistency and systematic, empirical checking and evaluation – ideally, in a continuous cycle of still more sophisticated knowledge and effective action.
Conclusions
Using the case of research into better and more learning conference, I presented the idea that a social-scientific theory can be a prescription for a better future, not just a description of the present. Such a transformative theory suggests that based on knowledge of the human potentials relevant to the domain (in our case, human learning), we may take certain actions to improve the domain. Because these action recommendations are derived from the theory and are eminently testable, they amount to what is known in classical scientific methodology as hypotheses. Once acted upon, the hypothesis claiming that if you do X in situation Y, Z will result, can be evaluated be checking if results X actually occur.  Of course, transformative research is about changing things for the better, not just changing them any which way. So the hypotheses are always a prediction that certain good results will ensue – otherwise, why bother take the action? Whether the produced consequences of the actions taken are indeed favorable is left for the stakeholders of the domain to assess, using whatever criteria are appropriate (user satisfaction, quality of life indices, health indicators, fairness of distribution, cost efficiency, etc.) and carefully guided and monitored by reflective researchers ensuring maximum validity, especially concerning the difficult questions as to what stakeholders are to be given what weight (but this problem is well-known by evaluation researchers). The main idea here is also the most difficult: that a theory is not necessarily about a current state of the world, but about a future desirable state. To understand this, we must switch from the classical philosophical view of knowledge as representational to a pragmatic view of knowledge as action-oriented (James, Dewey, Argyris, Schon). If knowledge is what helps us act intelligently in the world, and not just a set of static representations in our minds of an external world, a social-scientific theory could do just the same: help us act better in the world so as to create a desirable future for those involved – in a systematic and well-planned effort that is constantly monitored and evaluated experimentally, because that is what science is about. This pragmatic orientation in social research methodology could help alleviate the creeping irrelevance of the social sciences that has been pointed out by Bent Flyvbjerg (2001). He finds the mainstream sociologists’ attempts to create explanatory and predictive theories in the manner of the natural sciences to be misguided. Drawing on Aristotle, Foucault, Bourdieu and Dreyfus, he argues for a closer connection between knowledge and action and proposes a “phronetic social science” that will critically examine power and values in the socio-political domain.  My feeling is that Flyvbjerg is not radical enough. Analyses of the misuse of power are fine, and his recent international study of cost overruns in mega-projects is a fine contribution, too (Flyvbjerg, Bruzelius and Rothengatter, 2003). What practitioners such as clients need, however, is a way to avoid mega-project overruns, not just a tally of how often it happens. That is the usual coitus interruptus of social research: Describe serious problems and leave it at that. If Flyvbjerg and consorts had an generally valid, realistic theory about how to avoid cost overruns, complete with many testable hypotheses and concrete actions to be taken, he would have a fair chance of really making a difference. But this requires that he shifts his scientific focus from analyzing current abuses of power and values to theorizing what better ends power and values could used for - and how. Let me add that a much more radical view of the predicament of social science was voiced by a deputy director at the Danish Ministry of Science, Technology and Development, the agency that oversees and funds Danish scientific research. He simply stated: “If the social sciences don’t produce something of relevance, we will have to shut them down” (Bertramsen, 2004), words ominously reminiscent of the government-initiated closure and reorganization of the Department of Sociology at the University of Copenhagen some 15 years ago. That is an indication of how far stretched is the government’s patience with sociology and allied social sciences in Denmark. The need to retool social research so as to make it practically relevant is pressing.  While many social researchers are involved in practice-oriented, nitty-gritty research that has obvious implications for social policy and the welfare of citizens, basic and overall notions of what social theory is must be refashioned. Society needs the contributions that astute and disciplined social scientists could make if they used one of the many approaches that take an interest in immediate practical effects – action research, participative research, intervention research, developmental action inquiry, etc. It is to this bundle of practice-oriented research methods that I wish to add the present reflections on transformative theory.
Literature
Ackoff, Russell (1953): The Design of Social Research. Chicago: University of Chicago Press.Ackoff, Russell L. (1981): Creating the Corporate Future. New York: Wiley.Argyris, Chris (1982): Reasoning, Learning, and Action. Jossey-Bass.Argyris, Chris (1999): On Organizational Learning. 2nd. edn. London: Blackwell. Argyris, D., Putnam, R. and Mclain Smith, D. (1985): Action Science. San Francisco: Jossey-Bass.Baburoglu, Oguz, og Ib Ravn (1992): Normative Action Research. Organization Studies, 13(1): 19-34. Bertramsen, René Bugge (2004): (Remarks made during a presentation at the conference ”Tomorrow’s Knowledge Organizations”, arranged by Learning Lab Denmark, Copenhage, May 26, quoted from memory).Brown, John Seeley, and Paul Duguid (2000): The Social Life of Information. Boston, Mass.: Harvard Business School Press. Cook, Thomas D., and Donald t. Campbell (1979): Quasi-Experimentation: Design and Analysis Issues. Houghton-Mifflin.Dewey, John (1938): Experience and Education. New York: Collier BooksFlyvbjerg, Bent (2001): Making Social Science Matter: Why Social Inquiry Fails and How It Can Succeed Again. Cambridge: Cambridge University Press.Flyvbjerg, Bent, Nils Bruzelius and Werner Rothengatter (2003): Megaprojects and Risk. Cambridge: Cambridge University Press.Gergen, Kenneth J. (1982): Toward Transformation in Social Knowledge. New York: Springer-Verlag.Høilund, Peter (1992): Den forbudte retsfølelse. Om lov og moral. Munksgaard.James, William (1907): Pragmatism. New York: Longmans, Green and Co.Kolb, Alice, and David A. Kolb (2005): Learning Styles and Learning Spaces: Enhancing Experiential Learning in Higher Education. Academy of Management Learning and Education, 4(2): 193-212.Lewin, Kurt (1947): Frontiers in Group Dynamics. Human Relations, 1: 5-41, 143-153. Nowotny, Helga, Peter Scott and Michael Gibbons (2001): Re-Thinking Science: Knowledge and the Public in an Age of Uncertainty. Polity.Owen, Harrison (1997): OpenSpace Technology: A User's Guide. Berrett-Koehler.Paul, Ellen Frankel, Fred D. Miller and Jeffrey Paul (eds.) (1999): Human Flourishing. (A speical issues). Social Philosophy and Policy, vol. 16(1), Winter. Cambridge University Press. Pinker, S. 2002. The Blank Slate. New York: Viking.Ravn, Ib (1986): Idealization as the Common Element in Interactive Planning and Social Research Methodology. S3 Papers 86‚Äë05, Department of Social Systems Sciences, Universi­ty of Pennsylvania. Ravn, Ib (1989): “Implicate Order” and the Good Life: Applying David Bohm’s Ontology in the Human World. Ph.D. Dissertation, Department of Social Systems Sciences, University of Pennsylvania, Philadelphia, USA.Ravn, Ib (1991): What Should Guide Reality Construction? In Frederick Steier (ed.): Research and Reflexivity, pp. 96-116. London: Sage Publications. Ravn, Ib (2004): Action Knowledge in Intellectual Capital Statements: A Definition, a Design and a Case. International Journal of Learning and Intellectual Capital, 1(1): 61-71.Ravn, Ib (2005): The Learning Conference. To be presented at the Academy of Management’s Annual Meeting, Honolulu, Hawaii, USA, August 5-10. Ravn, Ib (forthcoming): Sociologi: Problematisere magtrelationer eller kvalificere sociale eksperimenter? Chapter 9 in Forskning i sammenhænge, Multivers, 2005.Reason, Peter, and Hillary Bradbury (eds.) (2001): Handbook of Action Research. London: Sage.Reason, Peter, and William R. Torbert (2001): The Action Turn: Toward a Transformational social science. Concepts and Transformations, 6(1): 1-37. Rogers, Carl R. (1961): On Becoming a Person. Boston: Houghton Mifflin.Rogers, Tony (2003): Conferences and Conventions: A Global Industry. London: Butterworth-Heinemann.Rothman, Jack & Thomas, Edwin J. (eds.) 1994. Intervention Research: Design and
Development for Human Service. New York: The Haworth Press.Schon, Donald (1983): The Reflective Practitioner. New York: Basic Books.Seligman, M.E., og M. Csikszentmihalyi (2000): Positive Psychology. An Introduction. American Psychologist, 55(1): 5-14.Sorang, Ole (2004): Personal communication (from the sales director at Radisson SAS).Susman, Gerald I., and Roger D. Evered (1978). An assessment of the scientific merits of action research. Administrative Science Quarterly, 23, 582-603.Vygotsky, L.S. 1978. Mind in Society. Boston, MA.: Harvard University Press.