Je ne crois pas aux choses mais aux relations entre les choses.
– Georges Braque

1Over the past 10 years, I have worked on the problems raised by the governing, the governance, and the governability of complex organizations and socio-economic-political systems. This terrain has been explored by many researchers. Indeed, over the last 10 years, these themes have become the centre of important debates on every continent (Kumon 1992; Kooiman 1993; Castells 1996). However, governance studies are still in their infancy. Although many interesting approaches and perspectives have been proposed, there is still no consensus on the best way to handle these issues, nor is there agreement on a lexicon or vocabulary for formulating these questions.

2This compendium of papers is a progress report on work that strives to generate new responses to these problems. These studies are unified by the choice of a particular strategy for dealing with governance problems. It is not intended to be le dernier mot on these matters, but only a premier effort to clarify the issues, using a method that has proved useful—the social learning approach.

CHANCE EVENTS AND INSTITUTIONAL SUPPORT

3The research program on which this volume is based originated in 1988-89 during a sabbatical leave from the University of Ottawa, but the original plans evolved as a consequence of many chance events and help from friends and colleagues at certain junctures.

4The first chance event was an invitation from Rod Dobell, president of the Institute for Research on Public Policy, to spend my sabbatical at the Institute. At the time, the Institute was very much a forum interested in exploratory thinking. My work there was entirely curiosity-oriented, but I had an opportunity to discuss the results of my first analyses with Peter Dobell, Rod Dobell, Jeffrey Holmes, Steven Rosell, and Walter Stewart. Each of these colleagues forced different perspectives on these issues of governance on me, and they all had an impact on the reflections that stemmed from my sojourn at the Institute.

5The second chance event was David Zussman’s invitation to present a paper on these issues at the Aylmer Conference of the Liberal Party of Canada in November 1991. This was an opportunity for a synthesis of work that, at that point, remained somewhat scattered and issue-oriented. Preparing a synthetic piece on the Strategic State was an occasion to refine my approach considerably and to sketch, however roughly, a research program that is still unfolding. A fragment of this paper was published in the proceedings of the conference early in 1992; a three-part document covering the same terrain much more thoroughly was published in Ciencia Ergo Sum (Paquet 1996a, 1997a,b). The latter version is included in this volume (Chapter 11).

6A third determining event was an invitation to become a senior research fellow at the Canadian Centre for Management Development (CCMD) in 1992. This opportunity, offered by Ralph Heintzman, who was then vice-principal (research) at CCMD, together with my extensive involvement with the teaching faculty at the Centre, at the request of Lise Pigeon, provided an extraordinarily rich cauldron in which these ideas could be stirred, debated, and further distilled. From 1993 on, various working papers were used in the classroom at the Centre; in them, the ideas were tested and polished through discussion with hundreds of senior executives in the federal public service. A number of papers emerged from this extremely rich experience at CCMD; a few have been collected here.

7The final important event was the creation of PRIME (Program of Research in International Management and Economy), which John de la Mothe and I pressed into existence in 1993 with the support of Dean Jean-Louis Malouin. This organized research unit brought together a number of colleagues (Robert de Cotret, Georges Hénault, Luc Juillet, David Large, Paul Laurent, Morris Miller, Christian Navarre, Jeffrey Roy, Robert Shepherd, Chris Wilson, and others) and a number of associates from the private, public, and civic sectors. The PRIME group has produced a stream of papers and books, in which we have all become more and more concerned with problems of governance.

8The success of the PRIME experience — both in terms of the interest it generated in the community and the financial support it elicited from the Social Sciences and Humanities Research Council, Industry Canada, Statistics Canada, etc. — prepared the way for our 1997 petition to the University of Ottawa for the creation of the Centre on Governance. The Centre is an umbrella organization for a variety of programs of research and organized research units with an interest in governance issues. The support of Vice-Rectors Howard Alper and Gilles Patry and of Deans Caroline Andrew (Social Sciences) and Jean-Louis Malouin (Administration) was most important in efforts to create the Centre and to assure sustained funding for its development phase.

9The purpose of the Centre is not only a certain mise en visibilité of the research personnel of the University of Ottawa and their partners, affiliates, and associates, but also, more importantly, the development of a particular manière de voir — an approach to governance issues that holds the promise of revealing important dimensions hitherto not accorded the attention they deserve, of leading to explorations of novel sorts of governance, and of providing opportunities to develop clinical interventions to ensure greater effectiveness for a number of organizations and sociotechnical systems.

UNE MANIÈRE DE VOIR

10Since the early 1970s, it has been my view that the overly simplistic and mechanistic ways of examining the problems of coordination and governance in modern socioeconomics are grossly inadequate (Paquet 1971). The triumphant belief — still in good currency in the 1960s — that the simple modeling of rational economic actors (with their target-and-instruments-type policymaking) could cure all ills had been all but discredited by the 1970s. But this view is still living on, even though the intervening years have revealed that most coordination and governance problems are “wicked” — the goals are ill-defined and uncertain and the means–ends relationships unstable and unreliable. It has become clear that neither the market mechanism nor rational policymakers (neither presumed perfect competition nor presumed perfect computation) can ensure that the socioeconomic system will be governed perfectly. An alternative way of looking at the coordination problem was needed.

11A major alternative to the mechanistic models of the 1960s and 70s was the systems approach, which emphasized the dynamics of interaction and interrelationships among actors. General systems theory, in particular, proposed a holistic way of thinking. Ludwig von Bertalanffy, a former faculty member at the University of Ottawa, was a most important voice preaching the systems approach gospel against the mechanistic stimulus-response view of the world that was in vogue. But his perspective was not taken very seriously. Indeed, many opinion molders regarded it as suspect because of its link to discredited theories of vitalism, which held that organisms and organizations were directed from within by a soul-like force (von Bertalanffy 1968; Davidson 1983).

12Yet, von Bertalanffy’s perspective was never vitalistic. It focused on organizations as open systems and on a reality best represented as a many-layered architecture of organizational entities. It searched for laws (progressive integration, differentiation, mechanization, centralization, etc.) that might apply to every layer of the system, from cell to biosphere. But this gambit was perhaps too ambitious, for, in the 1960s and 70s, the systems approach seems to have lacked sufficient heuristic power to shake off the seduction of the language of management science.

13The management science approach to governing presumed that public, private, and civic organizations were strongly directed by leaders who had a good understanding of their environment, of the future trends in the environment if nothing were done to modify it, of the inexorable rules of the game they had to put up with, and of the goals pursued by their own organization. Those were the days when the social sciences were still Newtonian and pretended to explore a world of deterministic, well-behaved mechanical processes where causality was simple because the whole was the sum of the parts. The coordination–governance challenge was relatively simple: building on the well-defined goals of the organization, it was to design the control mechanisms likely to get the organization where its leaders wanted it to be.

14Many issues were clearly amenable to this approach, and many still are. But as the pace of change accelerated and the issues grew more complex, private, public, and civic organizations were confronted more and more with “wicked problems” (Rittel and Webber 1973). In dealing with such problems, inquiry can only mean “thinking and acting that originates in and aims at resolving a situation of uncertainty, doubt and puzzlement” (Schon 1995: 82). This calls for a new way of thinking about governance. At best, one can hope for pattern causality: the gradual construction of a “causal” story on the basis of background knowledge of the system that is often tacit, and “working back,” as plumbers do when tracing a leak to its source. In this quantum world, there is no objective reality, the uncertainty principle looms large, events are at best probable, and the whole is a network of synergies and interactions that is quite different from the sum of the parts (Becker 1991).

15Three important forces have played a central role in generating this quantum world: the rise of the international, flexible production system, the accelerating pace of technological change, and the new global financial structure. As a result of these, governments and state authorities have lost much of their dominion over national economies and societies, and there has been a decline in state legitimacy (Morales 1994; Strange 1996).

16This erosion of the power and legitimacy of the state has had two important impacts: first it shifted attention to the nonstate authority, to the other loci or sources of power; and, second, it brought nonpurposive action and unintended consequences to the centre of the stage (Galston 1998). A number of important studies have explored these different sites of power and tracked down the ways in which much of the state authority has become diffused to nonstate agents in economy, polity, and society (Horsman and Marshall 1994; Held 1995; Strange 1996). This phenomenon has given rise to a new distributed and not entirely purposeful governance shared among the different stakeholders as the new emerging social technology.

The Boulding Triangle

17Even in the old world of governance, the boundaries between the economic, political, and civic spheres were never well-defined either conceptually or statistically; they did not correspond to a rigid frontier, but rather to a wavering and evolving fracture zone between subsets of organizations and institutions integrated by various mechanisms. This has become even more true in the new world of governance.

18Economists have explored this terrain for quite some time. François Perroux (1960) and Kenneth Boulding (1970) proposed a simple conceptual map. Both identified three generic ensembles of organizations dominated more or less by a different mechanism of integration: quid pro quo exchange (market economy), coercion (polity), and gift, solidarity; or reciprocity (community and society). These mechanisms had been explored by Karl Polanyi (1957) as dominant features of the concrete socioeconomics of the past. Perroux and Boulding fleshed out the idea and applied it to the modern context.

19In this approach (Figure 1), the organizational terrain is roughly divided into three domains where the rules or mechanisms of coordination are based on different principles: the economic/market domain (B) where the forces of supply and demand and price mechanisms are the norm; the state domain (C) where the rules are based on coercion and redistribution; and civil society (A) where cooperation, reciprocity, and solidarity are the integrating principles. This corresponds roughly to the partitioning of human organizations into economy, polity, and society (Wolfe 1989).

20A careful survey of many advanced sociopolitical economies reveals that society, economy, and polity each occupy approximately one-third of the terrain and that the central point is a rough approximation of the centre of gravity of the organizational triangle. This does not correspond to the statistical portrait emerging from official agencies, mainly because zone-A activities are grossly underreported; activities in the home, within not-for-profit associations, and in general beyond the market and the state are poorly recorded and remain largely underground (Paquet 1989a).

21These three sectors have not always had equal valence and need not have similar weight. A century ago in Canada, the state portion was quite limited and the scene was dominated by the other two sets of organizations. From the late 19th century to the 1970s, government grew in importance to the point where probably half of measured activities fell into the general ambit of state and state-related activities. The boundaries have been displaced accordingly over time. More recently, a vigorous counter-movement of privatization and deregulation has caused a reduction in the state sector and a reverse shift of the boundaries (Chapter 11).

22In parallel with these swings, there has been a tendency for the new socioeconomy to trigger the development of an ever-larger number of mixed institutions, blending the different mechanisms to some extent (market-based public regulation, public–private–social partnering, etc.) to provide the necessary signposts and orientation maps in a new confused world. In the recent past, this has translated into a much denser filling in of the Boulding triangle. Mixed institutions have been designed that are capable of providing the basis for cooperation, harmonization, concertation, and even co-decision-making involving agents or organizations from the three sectors (Leroy 1990; Burelle 1995; Laurent and Paquet 1998).

Heterarchy and Co-evolution

23A modification of the governance process necessitates some rearrangement of the role of each sector and, therefore, entails a shift of the boundaries between A, B, and C. Any such shift corresponds to a new division of labour among the three sectors, but there is not necessarily a hierarchy among those sectors.

24Indeed, the great weakness of most analyses of the scope of government has been that they ascribe to the state either a dependent and somewhat secondary role vis-à-vis the market or a domineering role vis-à-vis economy and society. In the first case, the state is required to attend to matters only when neither the market nor civil society is able to take care of it; in the second case, the state is imposing hegemonic constraints on the other sectors. Both these positions are misleading ideologic stands. In reality, the relationships among sectors are heterarchical: it is a world without a pecking order. Heterarchy introduces “strange loops” of authority “under conditions of time and place” very much like the “game of paper, rock, and scissors where paper covers rock, rock crushes scissors, and scissors cut paper” (Ogilvy 1986-87). Any sector may at times have dominion over the others; indeed, the three sectors co-evolve.

25The ecological concept of co-evolution is an apt way to synthesize the links among these three universes. Co-evolution in biology refers to an evolutionary process based on reciprocal responses of closely interacting species. Reference has been made to the co-evolution of the beaks of hummingbirds and the shape of the flowers they feed from. The concept can be generalized to encompass feedback processes among interacting systems (social, economic, political) going through a reciprocal process of change. The process of co-evolution becomes a form of organizational learning, that is, of joint learning and interadjustment of economy, society, and state (Norgaard 1984).

26The central characteristics of this jointly evolving process are resilience (the capacity for the economy–polity–society nexus to spring back undamaged from pressure or shock through some minor rearrangements that do not modify the nature of the overall system) and learning (the capacity to improve present performance as a result of experience through a redefinition of the organization’s objectives and a modification of behaviour and structures as a result of new circumstances). These governing relations are in creative tension (resilience calls for preservation, while learning means change) and must be balanced. This does not call for a rigid division of labour among the spheres, but rather, for a capacity to switch to a greater or lesser dependence on one family of integrative mechanisms or another as circumstances change.

27One may identify a variety of mixes of political, social, and economic mechanisms (and different modes of interaction among government, business, and society) in different parts of the world. The Anglo-American system (Canada, United States, United Kingdom, Australia, New Zealand, and South Africa) is prone to ascribing a dominant valence to the market mechanism, to the point of belittling the scope of state and civil society. Other parts of the world (Western Europe, Asia, etc.) have chosen to assign a much greater role to the state, but also to community, culture, citizenship, and social cohesion (Dahrendorf 1995).

28In this process of co-evolution, adjustments are not the result of the workings of some invisible hand. The state has an important role in maintaining healthy communication in the forum and workable competition in the market. It also has an important intelligence function if it is to act as catalyst in an innovative learning process (Wilensky 1967; Lundvall 1992).

“Glocalization” and Dispersion of Power

29To cope with a turbulent environment, organizations must use the environment strategically, in much the same way a surfer uses a wave: to learn faster, to adapt more quickly. This calls for noncentralization, for an expropriation of the power to steer held by the top managers in an organization. This is very different from a unilateral decentralization that can be rescinded. There must be constant negotiation and bargaining with partners. Managers must exploit all favourable environmental circumstances and the full complement of imagination and resourcefulness in the heart and mind of each team player; they must become team leaders in task-force-type projects, quasi-entrepreneurs capable of cautious suboptimizing in the face of a turbulent environment (Leblond and Paquet 1988).

30This sort of strategy calls for lighter, more horizontal and modular structures, for the creation of networks and informal clan-like rapports (Bressand et al. 1989). This is the case not only in the public sector; in the private sector, the “virtual corporation” and the “modular corporation” are now the new models of governance (Business Week 1993; Tully 1993).

31These new modularized private, public, and civic organizations cannot impose their views on their clients or citizens. The firm, very much like the state or civic organizations, must consult. Deliberation and negotiation are everywhere, moving away from goals and controls="true" and deeply into intelligence and innovation. A society based on participation, negotiation, and bargaining has more and more replaced one based on universal rights. The strategic organization has to become a broker, a negotiator, an animateur; and, in this network, a consultative and participative mode obtains among the socioeconomy, the firm, the state, and communities (Paquet 1992a, 1994a; Cassells 1996).

32All this triggers a paradoxical outcome that has been analyzed by Naisbitt (1994) and christened “glocalization” by Courchene (1995). As globalization proceeds, economic integration increases, and the component parts of the system become more numerous. The central question is how to organize for faster learning. And it would appear, according to Naisbitt, that the game of learning is going to generate more innovation if those components confronted with different local realities are empowered to make decisions on the spot. Thus, globalization has led to localization of decision-making, to empowerment, to the dispersion of power, and to a more distributed governance process.

DISTRIBUTED GOVERNANCE

33In times of change, organizations can only govern themselves by becoming capable of learning both what their goals are and the means to reach them as they proceed. This is done by tapping the knowledge and information that active citizens possess and getting them to invent ways out of the predicaments they are in.

34This leads to a more distributed governance that deprives the leader of his or her monopoly on directing the organization. For the organization to learn quickly, everyone must take part in the conversation and contribute each bit of knowledge and wisdom that he or she has that has a bearing on the issue (Paquet 1992a; Webber 1993; Piore 1995).

35Distributed governance does not mean only a process of dispersion of power toward localized decision-making within each sector. It also entails a dispersion of power over a wide variety of actors and groups within the Boulding triangle, because of the fact that the best learning experience in a context of rapid change can be brought about through decentralized and flexible teams woven by moral contracts and reciprocal obligations negotiated in the context of evolving partnerships (Nohria and Eccles 1992; de la Mothe and Paquet 1994).

A Triangle-wide Governance System

36Distributed governance is embedded in a set of organizations and institutions built on market forces, the state, and civil society. But it is most importantly nested in transverse links relating these three families of institutions and organizations and allowing them to be integrated into a sort of neural net.

37These transversal links neither echo the traditional, functional top-down organization nor the matrix form of organizations, where vertical–functional and horizontal–process rapports are supposedly keeping one another in check. Rather, in a transversal world, processes are dominant, and the reaction to external challenges is for the different stakeholders to coalesce laterally to create informal links and multifunctional teams capable of promoting faster and more effective learning (Tarondeau and Wright 1995).

38Under ideal circumstances, this multifunctional esprit de corps provides a most fertile ground for social learning. It is based on the existence of a social capital of trust, reasonableness, and mutual understanding that facilitates the debates and generates a sort of basic pragmatic ethic likely to promote interaction and synergies among the many potential partners in each of the three families of organizations. But this entails mobilization of all participants through a wide array of coordination maps and institutions all over the Boulding triangle, and this may prove much more difficult to realize than is usually presumed. Indeed, not all social learning is feed-forward in nature, and, consequently, the neural net arrangements may encompass only a portion of the Boulding space, may link the various components only loosely, and may also generate “low” learning.

39In these forums that cut across bureaucratic hierarchies and vertical lines of power, fraught with overlapping memberships, personal ties, temporary coalitions, and special-task organizations, “the organizational structure of the future is already being created by the most as well as the least powerful” within the new paradigm (Hine 1977). Indeed, to the extent that middle-range regional and transnational networks are cutting across the usual structures, the interactions distill, in an evolutionary way, an always imperfectly bounded network (Strange 1996).

Transversal Governance and Meso-innovation Systems

40Our exploration of the evolution of the governance process suggests that this new pattern tends to evolve in two directions: its centre of gravity shifts downward toward the subnational level with a pattern of power distributed more broadly along the supra- to infra-nation-state axis; and its area spans a broader terrain involving a larger number of institutions and coordinating maps from the economic, political, and civic sectors.

41The addition of a major component of associative governance to the more traditional state and market governance mechanisms triggers a major qualitative change. It introduces the network paradigm within the governance process (Cooke and Morgan 1993; Castells 1996, 1997), and this paradigm not only dominates the transactions of the social sector, but also permeates the operations of both the state and market sectors (Amin and Thrift 1995). For the network is not, as is usually assumed, a mixed form of organization existing halfway along a continuum ranging from market to hierarchy. Rather, it is a generic name for a third type of arrangement, built on very different integrating mechanisms: networks are consensus or inducement-oriented organizations and institutions (Kumon 1992; Acs et al. 1996).

42Networks have two sets of characteristics: those derived from their dominant logic (consensus and inducement-oriented systems) and those derived from their structure. The consensus dominant logic does not abolish power, but means that power is distributed. A central and critical feature of networks is the emphasis on voluntary adherence to norms. Although this voluntary adherence does not necessarily appear to generate constraints, per se, on the size of the organization, it is not always easy for a set of shared values to spread over massive disjointed transnational communities: free riding, high transaction costs, problems of accountability, etc., impose extra work. So the imperatives of leanness, agility, and flexibility have led many important multinationals to choose neither to manage their affairs as a global production engine nor as a fully decentralized system, but as a multitude of quasi-independent units working in a loose confederated structure (O’Toole and Bennis 1992; Handy 1992).

43The structural characteristics of the network nicely complement the collaborative and adaptive network intelligence (Kelly 1994: 189). The network externalities and spillovers are not spreading in a frictionless world; they cast much more of a local shadow than is usually presumed: “Space becomes ever more variegated, heterogeneous and finely textured in part because the processes of spatial reorganization... have the power to exploit relatively minute spatial differences to good effect” (Harvey 1988). So a network does not extend boundlessly; instead, it tends to crystallize around a unifying purpose, mobilizing independent members through voluntary links, around multiple leaders in overlapping and superimposed webs of solidarity. This underscores the importance of “regional business cultures” and the relative importance of networks of small and medium-sized enterprises as a source of new ideas (Putnam 1993; Lipnack and Stamps 1994).

44Reciprocity, based on voluntary adherence, generates lower costs of cooperation and, therefore, stimulates networking as social capital accumulates with trust. Not only do the networks generate social capital and wealth, they are also closely associated with a greater degree of progressivity in the economy, that is, with a higher degree of innovativeness and capacity to transform because networks cross boundaries. Indeed, boundary-crossing networks are likely to ignite considerable innovativeness because they provide an opportunity for reframing. In the face of placeless power in a globalized economy, seemingly powerless places, with their own communication code on a historically specific territory, are fitful terrains for local collaborative innovation networks (Acs et al. 1996).

Renaissance-style Interdependency

45In the transition period from the current nation-state-dominated era to the newly emerging era of distributed governance and transversal coordination, there will be a tendency for much devolution and decentralization of decision-making, i.e., for the meso-level units in polity, society, and economy to become prominent and for the rules of the game of the emergent order to be couched in informal terms. Moreover, the emergent properties of the new order (be it a public philosophy of subsidiarity or another set of workable guiding principles) are likely to remain relatively unpredictable (Ziman 1991; Norgaard 1994; Paquet 1993a, 1995).

46This multilayered structure is something very like a neural net of the kind found in a living brain: a layered system of many signal-processing units interacting in parallel within and between layers. This sort of system can learn (i.e., transform) in reaction to external stimuli and develop a capacity for pattern recognition and for adaptation through experience. Indeed, the resiliency of the neural net (in the brain or in an organization) is due to the redundancy of connections that allows the information flow to circumvent any hole or lesion.

47The new form of transversal coordination now in the making may not suffer as much as some fear from the loss of central control and the weakening of the national state imperium. A different sort of imperium, adapted to the age of networking, is emerging — reminiscent of the Roman empire under Hadrian, where the institutional order was a loose web of agreements made to ensure compatibility among open networks (Guéhenno 1993).

ADDITIONS TO THE LEXICON

48To analyze this drift in the governance process, it is important to develop a lexicon capable of differentiating the various aspects of the transformation. Borrowing to a large extent from systems analysis, we must first distinguish between, on the one hand, the organization as an open system and, on the other, its environment and recognize that there is continuing interaction between them.

49The environment of the organization may be characterized in broad terms by its texture. And the texture of the environment has an impact on the system itself. Depending on whether the environment is placid, randomized, clustered, disturbed reactive, or frankly turbulent, organizations will have to vary and transform themselves in different ways to cope effectively with these challenges (Emery and Trist 1965). The texture of the environment has become much more complex, diverse, and dynamic over the last while, and it has changed much more rapidly than it used to. Problems are more often than not the result of a confluence of different factors, knowledge is dispersed over many actors, and there is much uncertainty and dissent about the nature of the objectives pursued. This has created new pressures on the organization as an open system.

50The governing activities of any single actor have become rather ineffective. Purposeful action by stakeholders is likely to encounter resistance from highly organized groups in these complex, dynamic, and diverse environments. Unintended consequences, external economies and diseconomies, and feed-back of all sorts are likely to ensure that the intended outcomes will not be achieved. The interaction and composition effects, the coordination and “collibration” (co-equilibration) efforts, and the important degree of integration, differentiation, and hierarchization that are likely to materialize, are bound to generate a pattern of governance that may have little to do with the original plans.

51Indeed, the pattern of governance is likely to emerge rather than being crafted.

Governing and governance are subjected to a permanent process of mutual interaction. Actors who govern, or try to govern, also influence the governance structure.... Some (more powerful) actors have the possibility to rewrite some “rules of the game” but no one has complete control. There is always some intended and unintended change, which creates maneuvering space for actors willing to change the existing pattern. [Kooiman 1993: 258-259]

52We are entering an era where the governance process is a game without a master. This raises the question of whether such systems are governable. Governability is a measure of the organization’s capability to govern itself within the context of broader systems of which it is part, and the environment within which it is nested. Governability makes no sense in a static context; it corresponds to the organization’s capacity to transform, its capacity to modify its structure, its process, and even its substantive guidance mechanism and orientation. To ensure governability, some balance must be maintained between autonomy and responsibility. Moreover, there must be some match between the needs and potentialities of the required organizational resources for the governability dynamics to be viable. Finally, governability requires substantial equilibration between effectiveness and legitimacy (Kooiman 1993: 259-260).

53Governing, governance, and governability are obviously in continuous interaction: the gaps between governing needs and capabilities are likely to modify governing behaviour and transform the governance pattern. This is likely to trigger the emergence of a fitful degree of centralization, differentiation, and self-governance; to give rise to a variety of partnerships and joint ventures to respond to the challenges posed by knowledge dispersion, motivation, and implementation problems; and to correct some of the important side-effects of the existing governance structure.

54The emerging institutional order may not correspond to one ensuring optimal governability, for it is not determined on the sole basis of efficiency; the most important dimensions are legitimacy, fairness, ethics, learning, etc. The overall objective is to maintain enough coherence over time to maintain the organization as a bundle of coordination mechanisms, but not so much that it would prevent the organization from developing new instruments, new perspectives, and new purposes (Laurent and Paquet 1998).

55This subtle search for the right degree of coherence calls for a new political language to replace the traditional engineering language in the world of governance. The dynamic new realities of alliances, power, influences, and constituencies have replaced the old static realities of property, structure, planning, and control. In a world where the new assets are intangibles and mainly in the control of stakeholders, the challenge of governability is the challenge of transforming mercenaries, owing loyalty only to themselves, into members of a community interested in and capable of allegiance and reciprocal commitment (Handy 1998).

56This calls for significant modification in the form of our organizations. First, it confirms the need for modular and federal structures better able to mobilize loyalty. Second, it requires that trust be nurtured, as it must be in the bloodstream of the organization for it to be effective even though it is only loosely structured.

57This does not eliminate the responsibility of the state for “the infrastructure of life” (Handy 1998: 223) in the new governance. Without it, governability is in doubt. This new strategic role of the state is bound to be more modest than it has been in the last 50 years, but it is a most fundamental role in providing help to ensure that appropriate organizations can evolve, that citizens connect themselves better with the market, and that civic engagement and entrepreneurship are rekindled through permissive and supportive framework interventions (Handy 1998; Paquet 1999).

SOCIAL LEARNING

58In the dynamic, innovative, and cooperative environment of the learning economy, the capacity to learn increasingly determines the relative position of individuals, firms, and national systems. New modes of production of knowledge and new modes of collegiality, alliances, and sharing of knowledge have evolved (Gibbons et al. 1994; Lundvall and Johnson 1994).

59The learning economy is the source of wealth creation and is rooted in a social or collective mobilization of knowledge: learning is harnessing the collective intelligence of the team as a source of continuous improvement (Florida and Kenney 1993). This, in turn, commands a degree of cooperation to take advantage of positive externalities, economies of scale and scope, and strong cumulative experience–learning processes (Jacquemin 1995). But this process does not necessarily work perfectly.

60Although much know-what and know-why has been ever more effectively codified and can be produced and distributed as a quasi-commodity, know-how and know-who have remained tacitly and socially embedded (Foray and Lundvall 1996). Consequently, the production and distribution of these latter forms of knowledge have been more problematic; they depend a great deal on social cohesion and trust, on much trespassing and cross-fertilization among disciplines and on the development of networks capable of serving as two-way communication links between tacit and codified, private and shared knowledge, between passive efficiency-achieving learning and creative–destructive Schumpeterian learning (Boisot 1995). There are ample possibilities for coordination failures that can slow down the process of learning (de la Mothe and Paquet 1997).

Interaction and Conventions

61Interaction is necessary to generate effective learning. It focuses on the desirable form of imperfect competition or mixes of competition and cooperation characterized by product-based learning. Learning entails “the mutually consistent interpretation of information that is not fully codified, and hence not fully capable of being transmitted, understood, and utilized independently of the actual agents who are developing and using it” (Storper 1996: 259). It is of central importance because of the fact that knowledge is dispersed and exists in a form that is not fully codified. This calls for conventions or relational transactions to define mutually coherent expectations and common guideposts. These conventions differ from sector to sector: they provide the requisite coherence for a common context of interpretation and, for some, “cognitive routinization of relations between firms, their environments, and employees” (Storper 1996: 259).

62Such coherence results in nimbleness in the network economy. Yet, a good learning network must not be too coherent: the nodes should not be too similar nor the ties too strong or too routinized. This is the sense in which one may speak of “the strength of weak ties” (Granovetter 1973); a certain degree of heterogeneity and, therefore, social distance, might foster greater potential for innovation because the different parties bring a more complementary body of knowledge to the “conversation.” More fruitful synergies ensue.

Ideal-types of Transaction Structures

63To analyze the different types of transaction structures, Max Boisot (1995) has suggested a three-dimensional space — information space — which identifies an organizational system in terms of the degree of abstraction, codification, and diffusion of the flow of information within it (Figure 2). This three-dimensional space defines three continua: a vertical axis indicating increasing codification of the information (i.e., the more its form is clarified, stylized, and simplified); an eastward-pointing axis along which information is more widely diffused and shared; and a westward-pointing axis measuring increasing abstraction of the information (i.e., the more general the categories in use) (Boisot 1995).

64To illustrate the use of the information space, Boisot has identified several transaction structures corresponding to different loci. First, he identifies as “fiefdoms” the type of organization where information is very concrete and is neither much codified or widely diffused because of the fact that the core transactions are based on the savoir-faire and personal authority of the leader. Second, he characterizes the world of bureaucracy: information is more abstract (monthly reports, etc.) and codified in precise rules, but available only on a need-to-know basis within the organization and, therefore, not shared much. Third, is the market organization, where price information is very abstract, highly codified, and widely diffused. Finally, Boisot labels as “clans” some organizations based on shared values and personal contacts: in the clan, information is concrete, noncodified, and widely diffused.

65In each of these subspaces, governing and governance connote rather different realities. In fiefdoms and bureaucracies, governing is based on top-down command and governance is rather hierarchical; in markets and clans, governing is much more a lateral endeavour, and the pattern of governance much more horizontal and transversal.

The Learning Cycle and Learning Blockages

66Within the cube, Boisot has attempted to stylize the operations of the social learning cycle to capture the different phases of the processes of production and diffusion of information in organizational learning. This cycle is presented in two phases with three steps in each phase (Figure 3): phase I emphasizes the cognitive dimensions of the cycle, phase II the diffusion of the new information.

67Learning begins in phase I with some scanning of the environment and of the concrete information that is widely known and diffused (s) so as to detect anomalies and paradoxes. In step 2, one is led to stylize the problem (p) posed by the anomalies and paradoxes in a language of problem solution. The third step purports to generalize the solution of the specific issue to a broader family of problems through a process of abstraction (at).

68In phase II, the new knowledge is diffused (d) to a larger community of people or groups. Step 5 is a process of absorption (ar) of the new knowledge by the population and its assimilation so that it becomes part of the tacit stock of knowledge. In step 6, the new knowledge is not only absorbed but has an impact (i) on the practices and artefacts of the group or community.

69Boisot also notes the possibility of blockages at each step in the learning cycle. For example, in phase I, cognitive dissonance in s may prevent the anomalies from being noted, epistemic inhibitions of all sorts in p may stop the process of translation into a language of problem solution, blockages preventing the generalization of the new knowledge because of the problem definition being encapsulated within the hic et nunc (at) may keep the new knowledge from acquiring the most effective degree of generality. In phase II, the new knowledge may not be diffused appropriately because of property rights (d), certain values, or very strong dynamic conservatism which may generate a refusal to listen by those most likely to profit from the new knowledge (ar) or because of difficulties in finding ways to incorporate it (i).

70It is important to note that the social learning cycle does not pertain only to the search of new means to reach well-defined ends. It is double-looped in the sense that as the learning proceeds, anomalies and paradoxes are generating the redefinition not only of the means but also of the ends (Argyris and Schon 1978).

Interactivity

71Social learning is organization-based and interactive. It stems from creative interactivity. Interactivity is a form of dialectical relations among agents and their evolution through time. It connotes the process through which four aspects of organizations become harmonized: the various capabilities or competences (technical, organizational, strategic, learning) of organizations; the particular capacities of the different organizations (to solve problems, to absorb knowledge, to innovate and experiment, and to incorporate new knowledge in its functions); the interactions with the environment and with other organizations; and the degree of dynamic increasing returns for organizations in learning by learning.

72Interactivity brings some sort of cumulative process of learning built on externalities of all sorts, with great potential for irreversibility and inflexibilities of all sorts (Le Bas 1993: 13). But it mostly provokes the genesis of institutions: a set of guideposts, the locus for the memorization and transmission of routines and tacit knowledge through conventions, contracts, and contraptions that form a cognitive framework that guides the learning process and constrains the nature of the exploration and exploitation of new knowledge: they orient the directions of learning (March 1991; Llerena 1997).

73The nature of the ethos and of the culture can have an important impact on the shape of the learning cycle. One might find the learning cycle jammed in a narrow band close to the abstraction–codification plane on the left when learning is restricted only to a very limited community, or the learning cycle may rotate almost exclusively within the bureaucratic world or the market world, or it may be disjointed into several separate or quasi-separate loops within the Boisot cube.

74New types of relationships have developed in this new context: new open self-elective communities transcending borders and generating new bonds of a nonnational sort have emerged. But there has also been much stunting of the existing pluralistic relationships: the rise of reactive exclusionist “identity groups” defined by a total allegiance to a single club — be it tribe, race, gender, ethnicity — that can only lead to the politics of divisiveness and the prevention of the sort of ongoing conversation that leads to social learning (Piore 1995).

Design Rationality

75In this new fluid setting where precarious new associative relationships develop, not only is the citizen somewhat uprooted, but the whole process through which the sociopolitical system learns and gets transformed becomes ill-defined as it is pulled at the same time toward the supranational and the subnational levels, and coordination tends to become much more complex as it becomes based less on hierarchies and more on associative networks of cooperation (Paquet 1997c,d).

76These looser forms of coordination may rise organically, but not necessarily. It is not always easy for a set of shared values to spread over massive communities. There are organizational diseconomies of scale. This is why networks do not extend boundlessly, and why their development often depends on shocks (the moral equivalent of a war or the sociological equivalent of a defeat that sometimes provides the requisite esprit de corps) or on crises revealing a lack of trust, a lack of the requisite amount of social capital, and the erosion of communities.

77These looser forms of coordination build much on tacit knowledge, and tacit knowledge and incomplete knowledge create difficulties and botdenecks. But it can be argued that excessive codification may well also be a source of inertia and deceleration in the process of learning and change (Foray and Lundvall 1997). Mode 2 production of knowledge is the world of “delta knowledge,” i.e., of practical transdisciplinary knowledge as a result of reflection-in-action (Gilles and Paquet 1991). This sort of knowledge is not new, but it has become immensely more important in the recent past, as the intensity and complexity of interactions between actors in organizations has increased.

78The pressure to organize, learn, and innovate has generated the emergence of value-creating partner systems in which core competencies are often embodied in forms of knowledge that are idiosyncratically synergetic, i.e., in some form of connoisseurship or practical wisdom, or savoir-faire, that remains largely tacit but is fundamental nonetheless. This form of knowledge has been so neglected and the more traditional form of technical and codified knowledge has been so celebrated that very little has been done to uncover the ways in which delta knowledge is produced and diffused. Even if the capabilities on which this sort of connoisseurship is built are highly regarded, they are often considered as inimitable and noncontestable, so little is known about ways to augment them.

79The accumulation of reservoirs of tacit knowledge emerges from a conversation with the situation and from a process of exploration and learning driven by design rationality (Schon and Rein 1994). Such a process embraces error as the only way to learn, as the way to fuel creative deliberations. This process of learning through a conversation with the situation and as a result of errors (the difference between what is expected and what happens) is at the core of the learning organization, but it is also a quagmire that few have explored seriously. Consequently, we often count on the forces of instrumental rationality and logical processes only because of the fact that these are regarded as the only source of valuable knowledge. Only when the full array of different types of knowledge becomes legitimate and when we have probed the way in which nonlogical processes can be fully tapped can we hope to make the greatest and best use of mode 2 knowledge production.

PROPOS D’ÉTAPE

80The essays in this volume were written over some 10 years, and they do not form a systematic study of governance, as they lack the sustained argument of a monograph. However, the whole is more than the sum of the parts. Together, they convey an understanding that separately would be missing. This is the rationale behind my decision to publish them in this form.

81Part I contains two chapters that act in lieu of a fully developed conceptual framework. Chapter 1 develops more fully the argument presented in this introduction. Chapter 2 presents, in a succinct way, the social learning approach I have elected to use as our compass and makes passing references to a number of papers presented in part II as an illustration of the heuristic power of this approach.

82Part II is a collection of papers in which the social learning approach is used as the analytical framework that is most likely to serve as a set of useful organizing ideas.

• At the international level, this approach covers a vast terrain: a deconstruction of the free trade debate, and an outline of what a science and technology policy might be under free trade.
• At the national level, the social learning approach is used to study the aborted process from which an energy policy might have emerged, and the treacherous problems raised by the environment–energy interface.
• At the social level, the social learning approach is used to examine two thorny policy areas: multiculturalism and liberal education. These areas are ill-structured and poorly understood; my hope was, at best, to provide a provisional lay of the land.
• At the administrative level, I have examined two areas — public service commissions and granting councils — where extraordinary opportunities have been missed as a result of a systematic underestimation of the power of administrative mechanisms as a basis for reform.

83Part III attempts to draw some conclusions from these preliminary studies. It sketches the contours of the emerging strategic state, explains the importance of moral contracts in the new governance, and explores the way in which distributed governance and transversal leadership may materialize.

84The conclusion raises the question of the burden of office of citizens and officials alike in the new governance, examines the central issues of accountability and ethics at the core of social learning, and identifies the sort of connoisseurship necessary to survive in this world of 360-degree accountability.

85Finally, I would like to thank Vicki Bennett, Sandra Garland, and Marie Saumure for their help in getting this manuscript through the production process.