Introduction

In Simpson and Gill (2007) we noted that design is contingent and always open to re-interpretation as reality unfolds. For organizations, this reality is primarily socially constructed through communicative processes. As practitioners involved in the design of activities for complex social programs, the interlinked web of conversations that create our current reality is shaped by interaction across organizational, cultural, political and professional paradigm borders. We believe practitioners should immerse themselves in all those conversations and recognise the patterns that describe this emergent reality. It is essential that these conversations must be broadly inclusive if a holistic view of the design context is to be achieved. They must work with rather than retreat from the discursive realities of those communities with which the practitioner seeks to engage.

Our approach is founded in theory similar to that described by Norman Long as actor-oriented theory, which is “philosophically grounded in a social constructionist view of change and continuity” (Long, 2001: 2). This paper inquires into the privileged position of the practitioner as a particular type of actor within a complex social network, and specifically looks at how practitioners design their actions for intervention in complex social systems. The purpose of this is to provide a reflection on social system design practice as one side of a reflexive process of continued learning about practice.

To provide a theoretical basis for a practical generalized case study we have borrowed concepts from a range of theorists and disciplines, consistent with our view of practice as transdisciplinary. Our approach is also similar to Bourdieu’s view that ‘one cannot think well except in and through theoretically constructed empirical cases’ (Bourdieu, quoted in Webb, 2002: 48). Our method has been to review a number of theories of how social construction through conversation can be understood and then to use these theories as a way of viewing the problem of intervention design in practice.

Our theoretical themes are drawn from Long’s focus on situated social action, which in turn draws on the notion of lifeworlds, the lived-in world of the social actor (Long, 2001: 54). The theoretical basis we present is based primarily on the work of four people: John Searle (2005), Jeffrey Ford (1999), Jurgen Habermas (1979), and Patricia Shaw (2002). In discussing Searle’s work, Habermas describes the “layer of worldview knowledge functioning in everyday life as the background with which the hearer has to be familiar … to … understand the literal meaning of speech acts and to act communicatively” and concludes that “communicative action provides the medium for the reproduction of lifeworlds” (Habermas, 1984: 337). Ford and Shaw, as practitioners in organizational intervention, show how communicative action operates in the conversations that constitute the daily discourse of our work. These four writers, coming from diverse disciplinary areas, all provide insights into how a social constructivist perspective, applied through communicative processes, can help practitioners re-vision their role in designing social system interventions.

After providing an overview of these theories, we present a generic case study and examine how these theories can usefully explain the role of the practitioner and help improve the way practitioners work. Finally, we provide our views on how conversation theory as a practical application of social construction in organizations might provide a better perspective on the process of design of social system interventions.

The practice of change as shifting conversations

Institutions as socially constructed artefacts

In this section we investigate a theory of how institutions are constituted, drawing on recent work by John Searle (2005). Searle’s definition of an institution is any collectively accepted system of rules, procedures or practices that enable us to create institutional facts, a special type of social fact. Searle defines three social primitives which he believes are the processes on which social and institutional facts are built: collective intentionality, assignment of function, and constitutive rules and procedures.

He defines intentionality as the feature of minds by which mental states are directed at or about objects or states of affairs. While this includes intention it also includes such things as beliefs, hopes, emotions, and perceptions. A social fact is “any fact involving the collective intentionality of two or more agents” (Searle, 2005: 6).

Searle observes that we “impose functions on objects where the object does not have the function … intrinsically but only in virtue of the assignment of function. Tools are the obvious case” (Searle, 2005: 7). Groups can collectively assign functions by assigning a status to that person or thing which then allows the person or thing to perform a function which could not be performed without the group’s acceptance of that status. For example, a knife has a physical structure that suggests its function while a coin’s function only exists because of collective assignment of function. Groups create these status functions, and human institutions are matters of status functions.

The third primitive that Searle considers as elemental in constructing institutional reality is the constitutive rules and procedures that produce social facts through collective assignment. These processes are part of an immensely extensive web of agreements that make up social reality. In Searle’s words “when the procedure or practice of counting X as Y becomes regularized it becomes a rule. And rules of the form X counts as Y in C are then constitutive of institutional structures”. He concludes that “the institutional ontology of human civilization … is a matter of status functions imposed according to constitutive rules and procedures” (Searle, 2005: 9). Searle thus sees institutional reality as a network of interlocking socially constructed facts, and it is these facts that Habermas sees as the layer of worldview knowledge that forms the background to everyday life.

In institutional matters in particular this background layer of worldview knowledge involves matters of power. In Searle’s view “Human institutions are, above all, enabling, because they create power, but it is a special kind of power. It is the power that is marked by such terms as: rights, duties, obligations, authorizations, permissions, empowerments, requirements, and certifications” (Searle, 2005: 10). Status functions can become vehicles of power in society, and social acceptance of status function involves acceptance of obligations, rights, responsibilities, duties, entitlements, authorities, and permissions, among many other forms of power relationships. In Searle’s view, society consists of a series of deontic power relations, which provide reasons for action, and make possible desire-independent reasons for actions (Searle, 2005: 10). A practitioner’s position as an expert creates just such a deontic power relationship.

Searle sees language as the fundamental social institution. Language gives people the power to represent, whether that be both a correct and truthful representation, or incorrect, or false. Institutional reality can only exist in so far as it is represented as existing. “A status function must be represented as existing in order to exist at all, and language or symbolism of some kind provides the means of representation” (Searle, 2005: 12). Language also allows the deontic powers of institutional facts to continue after their initial construction.

Conversations in social construction

Ford sees organizational change as a process driven through shifting conversations and compares his constructivist view against a more traditional structural-functionalist perspective (Ford, 1999). Constructivist “realities emerge and are maintained in and through conversations” ( Ford, 1999: 483). By conversation, Ford is referring to the total discourse in and around an organization, including both verbal and non-verbal communicative forms. In addition to the many explicit conversations in which they engage, for each person there is a complex web of background conversations that surround and infuse our explicit conversations as “taken for granted familiarity or obviousness that pervades our situation” (ibid.: 484).

Conversations are the process through which we construct reality as well as being the product of that construction. Ford describes the “intertextuality of conversations” (ibid.: 485) where all the linguistic products of our constructed reality in turn “create what is described, reported, explained, understood, etc” (ibid.: 485), which echoes Searle’s network of interlocking socially constructed facts and Habermas’s layer of worldview knowledge. In this way we can say that, “organizations are networks of conversations rather than have networks of conversations” (ibid.: 485).

In talking of change management, Ford describes conversation as the focus and unit of work. In this perspective, the work of an intervention designer requires an understanding of how to shift conversations to create commitments, and to act to accomplish those commitments. Ford points out that a monolithic view of change is therefore problematic. If change is shifting conversations, seeing change as removal of a problem (a standard approach in a structural-functionalist perspective) is a gross reduction of the scale and complexity of the work required.

Ford mentions “language shifts” using the example of a migrant language dying out as the dominant language takes over. In organizations, it is more likely that a new minority language will gain currency and grow to replace an older stable language. Sometimes this new minority language does not gain ascendancy and sometimes it is deliberately stifled or ridiculed. Power and persuasion play an important role in whether new minority languages grow to replace older majority languages. Change is often the result of introducing a new language but this in itself is not sufficient to cause change. The new language must be an innovation that aids sense making.

Communicative rationality

Drawing from Searle’s ideas on language and conversation, Habermas (1979) advocated the need for ‘propositionally differentiated and institutionally unbound speech actions’. This translates into the need to avoid systematically distorted communication; wherein privileged world views are imposed over others (possibly legitimized through, for example, institutional power relationships). The key to breaking down these distortions is to be reflexive on the frames of meaning that underpin our ‘speech acts’. Conversation must be free to move to increasing levels of reflection; there must be freedom to check questioned claims, to evaluate explanations, modify a given conceptual framework, and to reflect on the nature of knowledge. Prohibitions or commands must be open for discussion, where they can no longer be taken for granted, justifications are assessed, and conceptions of norms are open to reflection as is the nature of prevailing political will. This is the character of ‘institutionally unbound’ communication, or of an ‘ideal speech situation’ where participants have an equal opportunity for discussion, are free from all domination, and are willing and able to reflect on the layered backgrounds of their worldviews.

According to Habermas (1984), communicative process involves the raising of ‘validity’ or ‘truth’ claims. These claims, when contested, can only really be resolved through conversation. When we detect a truth claim we tend to take a stance towards it and would naturally seek to test its validity through discussion. Communicative rationality refers to our capacity to engage in argumentation through which to resolve these claims. This argumentation should be unimpeded by social and political arrangements that would preclude discussions or conversations of this kind. Habermas provided an account of a form of deliberative process appropriate to the complexity of modern society. Here the emphasis is on discussion wherein participants develop a level of general agreement based on how people perceive what is right for the group rather than just on what is right for them. The presupposition is that people will be able to work out what is good for the group, or community that they are a part of by working as a group through discussion rather than individually.

The main point about all this for the design practitioner is that the fundamental values and understandings that frame complex social system intervention design may well be quite different if developed through a deliberative process rather than being constituted through the privileging of one theory-derived position or another as the focus for that development activity. Purposefully enabled conversation is the key.

The practitioner’s role in conversations

We can compare the perspective where a speaker is seen as a reporter of a true reality against the perspective where a speaker is constructing reality. In the second perspective, the speaker is responsible for the reality they construct. When we recognise our responsibility for the reality we construct, we must also examine our practices and how our words impact on others.

Shaw (2002) sees organizations as social realities created as a mix of both intended and unintended consequences of planned and unplanned actions; organizations are constituted as paradoxical phenomenon that can be at the same time both ordered and disordered, stable and unstable, or organized and disorganised. Shaw describes a way of seeing organizations in terms of “self-organizing patterning of communicative actions in complex responsive processes of human relating” (ibid.: 20).

Shaw’s views on practice are particularly relevant to social system intervention design. Shaw talks about a change in rationality, “from a thought before action, design before implementation, systematic, instrumental logic of organizing, towards a paradoxical kind of logic in which we see ourselves as participating in the self-organizing emergence of meaningful activity from within our disorderly open-ended responsiveness to one another (ibid.: 30).

This has significant implications for practice and for those practitioners who are “people seeking to be instrumental in shaping the world according to their expressed intentions” (Shaw, 2002: 30). In Shaw’s view it “is becoming increasingly clear that simple control over the outcome of complex interaction is indeed illusory” (ibid.: 30).

A common view of the practitioner is as a facilitator, a person that may be seen as helping “what was trying to happen to happen, and then get out of the way” (Shaw, 2002: 6). Shaw argues strongly for an alternative practitioner’s role as an integral and equal partner in the changing conversations. She describes this role as facilitative leadership and argues that it includes:

[A] practical feel for the process of shaping and patterning in communication … a keen sense of the move towards and away from agreement, of shifts in power difference, the development and collapse of tensions, the variations in engagement, the different qualities of silence, the rhetorical ploys, the repetition of familiar turns of phrase or image, the glimpsing and losing of possibility, the ebb and flow of feeling tone, the dance of mutual constraint … participating in the conversation in a way that helps to hold open the interplay of sense-making rather longer than would occur … to hold open the experience of not-knowing … to help people sustain an open-ended exploration and begin to notice the way they are generating useful ways of knowing and acting together as they do so … to shift people’s perspective to see that organizational change is this process rather than an end product of it... (Shaw, 2002: 33).

It is not too difficult to detect a strong correlation with Habermas’s recommended communicative rationality here. Through the emergent, discursively embracing conversation settings that Shaw is advocating, a group may develop insight that detunes the privileging of individual positions and ways of knowing that more conventional practice might have supported.

Responsibility for change through conversation

The conversational approach also requires participants to accept responsibility for the reality they are creating, calling for facilitative leadership in conversation. This introduces a deontological dimension to design which requires openness about intention, and explicit recognition of power and authority structures that might constrain a free and open conversation. Conversations in organizations can be both beneficial and harmful and are often at work through shadow systems. These are the “complex web of interactions in which social, covert political and psycho-dynamic systems coexist in tension with the legitimate system” that are “capable of spontaneous novelty and emergent strategy” (Shaw, 1997: 235). The “social, political and psycho-dynamic processes at play in the shadow system are as potentially destructive as they are creative” (Shaw, 1997: 246). The designer who knows the role of conversations in shaping future reality and who has developed skills in facilitative leadership has also gained power over those who do not have such knowledge and skills.

However, we should not overlook the challenges to facilitating conversations of this kind. We need to remember that people vary in their deliberative and rhetorical skills. While the settings for formal authority to speak may be in place, the ability of individuals to exercise this right will vary. We would recommend attention to communicative process or the importance of choices in relation to facilitation process. The need is to ensure a conversational setting wherein participants are as ‘equally privileged’ as possible. One possibility is to employ some form of cognitive mapping as a facilitation device (e.g., Gill, 2006). The use of a universal ‘cause and effect mudmapping’ language through which to manage a conversation tends to level the ‘linguistic playing field’ with reasonable success.

A design example

The preceding discussion has highlighted a fundamental need for reflexivity amongst all those who are empowered to influence the affairs of others. The call is for those with positions privileged by prevailing governance arrangements to admit a participative, dialogue-based foundation for activity design; to underpin the articulation of policy; and through which to develop new organizational structures. The development practitioner, for example, should strive to ensure that all organizational or governance arrangements are as consistent with the prospects for institutionally unbound conversation as possible. This way, the prospect for enhanced insight, collective ownership and sustained progress is maximized.

As a practical way of seeing the challenges inherent in this process, we’ve constructed a generalized case study of an intervention design, conceptualized as the design of a development project in a developing country. While we do not have the space here to provide a fully detailed case study, we offer a ‘skeleton’ (or general principles account) as the setting for a recommended conversation-driven design framework. That framework, we think, provides an alternative way of understanding the work of activity design that is quite different from a structural — functional – scientific approach. The conversation-driven design framework explicitly admits awareness of the complexity of socially constructed reality. It moves away from a design—implement sequential process to a process where design is intention that surfaces as part of every conversation at every stage of an activity. Design and implementation become intention and doing, and are always operating in tandem from the simplest conversation, to the total project level, and beyond to the level of organizations supporting the activity. There is a fractal-like reality to this view of design: the deeper we dig into the detail of an activity the more intentionality we find. Each and every conversation is part of the intention to change, and conversations that can best approach that status of being institutionally unbound are part of the design.

The design context

To illustrate the web of conversations active in creating a design context we will describe a generalized example based on a recent design activity. We have generalized this example to maintain the confidentiality of those involved, but also because it creates a story with themes that will resonate with many design practitioners. Many case studies, if generalized, would create a similar story.

The design study involved preparation of a plan for a set of activities. The main organizations involved in the story are a funding agency (‘the Agency’), their counterparts at the national level (‘National Counterpart’), a provincial coordinating bodies (‘Provincial Counterpart’), a provincial department responsible for reduction of poverty (‘Provincial Poverty Bureau’), and the local government. At the local government level, a Project Office had been established. This office trained township and village level officials in a planning and implementation methodology based on Participatory Rural Appraisal (PRA), which involved a self identification by village people of their need for assistance, and development of a program to provide this assistance in kind and through capacity building activities, mainly training and demonstration. Training and demonstration activities were provided primarily through technical bureaux, whose involvement was co-opted as required. Structurally, there were the normal coordination committees at every level and a set of project documents describing implementation plans and reporting on progress.

The Agency received a request from the local government to provide funding support and extend an activity that was soon to be completed under an existing agreement. The design work involved an assessment of the request for additional support and if considered feasible to prepare a detailed design document on which implementation could be based.

At the core of the existing activity was a participatory planning and implementation process at the village level. This process aimed to address the relationship between poverty and environmental degradation through increasing knowledge of the relationship between certain farming practices and hillside degradation and to introduce options for farmers that provided alternative sources of income while also rehabilitating degraded areas. This process was carefully implemented to ensure that the activity’s participatory principles were maintained based on respect for farmers’ local knowledge and the risk perceptions involved in changing well established farming practices.

The context of this fairly simple design task was not as simple. The design was to be set against Agency guidelines and needed to align with Agency policy. The Agency was in the process of adopting a new country strategy as well as new sectoral strategies. While local government had been implementing current activities under the existing strategies, new activities would need to be implemented under new strategies.

The National Government had also recently announced a major new policy to deal with issues associated with increasing income disparities between rural and urban populations and a perceived need to increase the efficiency of rural production. The new national policy increased the responsibility of local governments to provide vocational training for rural labour, both in terms of increasing on farm skills and also in terms of providing relevant training for farmers migrating to cities to seek paid employment.

Dominant discourses

Behind the conversations undertaken as part of the design mission were a number of dominant discourses and the degree of dominance of those discourses depended on an individual’s degree of commitment to their organization and its culture. Such discourses provide the ebb and flow of discussion about concerns of interest to the organization and its members. Again, behind the dominant discourses was the institutional and social reality of the participating organizations, the artefacts of previous and historical discourses.

Institutional reality is evidenced by the rules and guidelines associated with being part of any one of these organizations. The Agency had codified approaches to design developed over many years and many internal conversations. These conversations were in themselves constituted on the basis of other conversations about accountability and control. The codification of these concerns created a framework of guidelines that could be seen on the one hand as guidance as to what should be done and on the other hand in terms of limits on acceptability of work outputs. Seen positively these guidelines maintain standards, but seen negatively they limit innovation.

Social realities are evidenced in the different ways of talking and thinking at both the individual and group level. Cultural differences between the two countries involved are the most obvious example and are evidenced in different languages and different world views. But individually there are also differing world views and different life experiences, and these worldviews in part reflect age, gender and life experience as well as a acceptance or otherwise by the individual of the dominant discourse of their society, community or workplace.

During the first stage of the project the Agency had engaged a consultant to introduce a participatory planning and implementation system based on a PRA approach. In so doing, the Agency introduced a discourse which was well developed within the development community. Now approaching dominant narrative status, there could be some justification for fearing that its dominance would lead to a tyranny of idea and method (Cooke & Kothari, 2001). The perceived need for the project was driven by observations of a vicious cycle of poverty and environmental degradation, which itself was the construct of another expansive discourse within both development and scientific communities. Through conversations between the consultants and the County-level counterparts these discourses had been introduced into the Project Office and were evident in the way project staff spoke about their work.

In the Agency, the development of a new country and sector strategy was driven by broader discourses about the changing relationship between the National Government and the Agency and the need to improve efficiency of aid delivery, and to provide a greater impact by targeting a higher level of counterpart involvement. These discourses had influenced and constituted the Agency’s strategies, which in turn introduced elements of the Agency’s discourses into the design process through the requirement to make the design consistent with the new strategies.

A major discourse within the National Government was the movement of rural workers to the city. Televised discussions between economists about the best policies to deal with this phenomenon were evidence of an on-going and evolving discourse. Positive benefits of this process were seen as a reduction in pressure on land and the environment, potential increases in efficiency of production, and an increasing urban population to support growth in the exporting manufacturing sector. Problems were perceived in terms of social instability as families were separated and communities broken up. The National Government’s policy initiative was seen as a way of encouraging farmers to leave their land by providing skills that would increase their incomes in the city and make it easier to make a permanent shift to the city. For those that stayed behind, the policy provided for training to increase productivity and so create a more equitable distribution of incomes between the city and rural areas. The policy contained targets but left the details of implementation up to the Provincial and County levels of government.

The design task

The design task can be viewed from a traditional structural — functional perspective and from a communicative design perspective. In the first perspective the task involves basically a scientific exercise: observe and analyze the problem in its context, develop a theory that explains why the problem occurs, from that theory predict the outcome of a series of activities aimed at removing the problem, and document all this in terms of a logical framework of activities, outputs and expected achievements.

Standard methodologies for design focus on solving problems. A problem is identified and by removing the problem the situation is improved. This type of logic is well suited to the structural-functional perspective where problems can be seen as structural aberrations or functional blockages. When organizations are seen as a web of conversations it is more difficult to identify a single dominant problem.

When we look at the simple design problem above, the major discourses involve new policies in the Agency and the National Government. Set in and developing from these major discourses were the cascades of subsequent conversations within the Agency, within the National and Provincial Governments, and at the local government level. These conversations were interlinked with existing conversations within each organization, and are subject to differing rules in differing organizations. The conversations during the design period expressed different objectives and expectations, supporting and sometimes challenging existing organizational norms, and in so doing both defining and constituting the specific organizational culture of each organization. These conversations also aggregated and expressed the personal conversations which involve individuals with their families, their community, their peers, and their cultural heritage.

The original design problem was framed in terms of replacing degradation causing land use of steep slopes with rehabilitating land uses. Farmers are typically risk averse, so we need to demonstrate better means of providing a livelihood, with that advice based on incontrovertible scientific data. In this setting, it is often considered that a rational farmer must change his behavior when faced with “the facts”. The process is straight forward. Work closely with a small group of farmers and support them while they adopt the new technology. When their situation has improved and the advantages are clear, bring it other farmers to see how it has been done. We can use adult learning techniques to aid the uptake of technology. We will choose to work in the villages where demonstrations are most likely to be successful.

The community development specialist enters the design process bringing a different conversation. This conversation is about rights and giving people a voice. The interaction with farmers should not presuppose any outcomes, the farmers need to participate in the process, we need to understand their lives and we need to build our program around their choices. We will choose to work where there is the greatest need, where poverty is most entrenched.

The local government technical staff come into the design conversation with the weight of national targets bearing down on them. They need to choose villages consistent with national guidelines. Although the Agency can only support a certain number of villages in the extended activity, the national targets require them to work in twice that many. National policy dictates a limited range of technical interventions none of which are considered by the technical expert to be rehabilitation land uses.

Other national government agencies are working in the same area building roads and providing water supplies. Can we integrate efforts by working in the same villages? Apparently their organizational conversations are not open to us, there is no way in. Their targets are different and their policies do not require coordination with other departments. There is no room for an integrating conversation in which we can engage.

A structural-functional approach will not explicitly recognise these conversations and the incompatibilities between them. It may identify the problems as risks or maybe even ignore them by theorizing an implementation context in which they are assumed away. A communicative design will recognise these conversations and explicitly set about weaving them together, creating new conversations about the incompatibilities, helping people to recognise the background conversations in each discussions, bringing them forward so that constraints to agreement are identified and resolved. A conversational approach will recognise the social and institutional facts, acknowledge the likely history of their construction, and knowingly proceed with constructing a new social reality. When conversations are facilitated across the jurisdictions of individual organizations or across levels in the development hierarchy, they need to be conducted as far away from the dominant narratives of any single organization as is possible. This enhances our prospects for promulgating institutionally unbound conversation.

Some practitioners already do this knowingly; others do it naturally by being reflexive and open to the changing context of their conversations. However, the majority still work in a structural functional world where they imagine their work is so logically compelling that it will change the structures and functions around them.

Observations on the case study

A communicative design process should be explicitly cognisant of and be informed by the nature of those social systems within which it will operate. These social systems are constituted by formal procedures and policies as well as organizational norms and customs. Developments are both influenced by and will influence these associated social systems; quite complex feedback links may preclude precise understandings of how these links will play out in any planning exercise. To even further challenge ambitions for predictive planning traction, the social system domain may simultaneously connect organizational, industry, regional, national and international dimensions. This can be seen as similar to the state that complex system theorists describe as far from equilibrium.

Putting all this another way, we might suggest that the development field is one that is informed by ‘socially constructed’ concepts. The issues that development addresses, the multitude of attitudes that drive development agendas, advice and theories that drive the way we do and conceive development planning have all emerged through a process of social interaction. Naturally, the territory of social construction is contentious and excites many who would feel compelled to proclaim, without going any further, that issues like poverty, environmental damage and illiteracy (to give examples) are not socially constructed; they are real.

We suggest that these matrices of understanding that frame the concepts that development addresses (as partially illustrated by our case study) are idiosyncratically shaped by the particular experiences of individuals, of organizations, of nations and by professional disciplinary backgrounds. Different people, groups and organizations will have different things in their respective matrices. They will interpret development from differently constructed perceptions. When one person describes a particular issue, he or she needs to know that this issue may be quite differently interpreted by someone else. The ‘ideas’ we have about the issue will be different, possibly subtly, or possibly radically. This conceptualization is important, we think, because it flags our need to explore the diversity of meanings that might underpin the goals of our plans. Plans go astray when we set up or institute one planning response that does not fit the understandings of others with whom we will be interacting (for example, those people we are actually seeking to help) in relation to what they think is important and what should be done. We suggest that one pragmatic pathway forward is to embed a systematic and very intentional conversation dimension in our planning efforts. We need to engage with a community of discursive viewpoints. We need to note the diversity of viewpoints and understandings and we need to frame all our subsequent design efforts by the learning process that we facilitate through such an engagement process.

While many practitioners will automatically claim, at this point, that they do indeed engage in conversations and discussion; that they do know that different people have different views; that they would never impose solutions that do not fit; we would simply note that ‘there are conversations, and then there are conversations’. There is much in the way of guidance and insight to be derived from the philosophical domain that might sharpen our efforts and enhance our prospects to make even more of a difference. This paper has explored the prospects for some highly selective insights from the ‘communicative action’ domain to add value to development design and catalyze change that most would think to be highly desirable. We can only address a minute fraction of the insight that can be derived from the ‘applied philosophical domain’. We hope that what we do bring forth here is found to be of pragmatic relevance.

In a previous paper (Simpson & Gill, 2007) we questioned whether projects could successfully implement participatory methodologies from within the rigid organizational settings that many development organizations embed. As discussed in this paper, effective communicative design requires governance arrangements that are flexibly adaptive to those complex social, environmental and economic settings that always overlap to confound the precise predictability of designed interventions. This call for flexible and adaptive organizational settings is even more critical when projects seek to engage across a diversity of organizations, each with their own distinctive ‘ways of doing and seeing things’.

The idea of engaging across a diversity of organizations suggests the potential of critical systems theory, and in particular the centrality of boundaries in critical systems theory, as a guide to practice (Midgley, 2000). Midgley et al. (1998) reporting on the development of critical systems through the works of Churchman, Ulrich and Midgley, considers that:

  • “The boundaries of analysis are crucial in determining how improvement will be defined during a systems intervention”;

  • “Pushing out the boundaries to make a systems intervention more inclusive may well involve sweeping in new stakeholders”;

  • There is a “need to rationally justify the setting of boundaries”, and;

  • (Following Habermas) “rationality is dialogical” and boundaries “should be defined through dialogue by all those involved in and affected by the intervention”;

  • “Some stakeholders and issues may be stigmatised by” the setting of boundaries “resulting in their potential marginalization during interventions,” and;

  • “It will sometimes be necessary to challenge a consensus on boundaries by seeking the involvement of people who might not be defined as directly affected or involved, but who may nevertheless have an important perspective to bring to bear on the boundaries of the intervention” (Midgley et al., 1998: 470).

The potential of critical systems theory is enhanced when considered alongside complexity theory. McGillivray studied a range of views on boundaries in order to “help organizational practice inform complexity theory, just as complexity theory informs practice” (McGillivray, 2006: 102). McGillivray’s work examined a substantial number of studies across a range of disciplines. She pointed out that the problems under consideration involved defining boundaries around sets of resources and that “these resources ranged from tangible business assets through personal learning to biodiversity” (McGillivray, 2006: 102). We believe that this is an important insight into the concept of boundaries between systems. It is clear that boundaries need to be conceptualized in terms of specific domains. The same actor included within one boundary in one domain may be excluded by a boundary in another domain. The linkage provided by each actor into a range of system domains drives the complexity inherent in social systems and increases both the importance and impact of communicative action in constructing complex social realities.

Final remarks

There is a growing theory associated with social construction that is highly relevant to the design of interventions into complex social systems. Whether it is a formal meeting or a chat over coffee, a conversation is active in creating the future. If design is also seen as an activity that is concerned with creating the future it is clear that conversation and design, if not the same thing, are at least part of the same process. For the practitioner this opens up a whole new set of roles and responsibilities. Everyone is now a designer. The technical expert should appreciate theories of social construction and develop appropriate communication skills. Everyone in a joint activity needs to understand the recursive, almost fractal, aspect of conversations. A single speech act inside a conversation, inside a project, inside an institutional context is both subject to and also constitutive of a future reality. In this context, collective communicative work involves wielding and being subject to Searle’s deontic power. The design practitioner is working in a space where meaning, reality and ethics are part of the job. We need to move further toward the deliberative democratic outcomes that institutionally unbound conversations can afford by approaching the potential Habermas identifies for ‘pure communication’ as the outcome of his advocated communicative rationality.