Leading through conflict means believing in the possibility of what does not yet exist. It requires focusing on the luminous opportunity that lies at the end of the tunnel of obstacles. Mark Gerzon
The premise: conflict is designed, not accidental
Every conflict specialist practitioner has stood in the wreckage of a relationship, a board, or a coalition and asked the same question: how did it come to this? The instinctive answer is to look for a villain, a scapegoat - a difficult personality, a bad actor, a moment of betrayal. But the more honest answer, the one that emerges after enough post-mortems, is that most destructive conflict was never caused by a single person. It was designed in.
A building does not collapse because the bricks were malicious. It collapses because the architect failed to account for load, for movement, for the stresses that every structure must carry. The same is true of the social and organisational systems we inhabit. A dysfunctional board, a fractured executive team, a coalition government that tears itself apart, these are not primarily failures of character or evidence of malice. They are failures of conflict design. The structure was built in a way that made destructive conflict more likely than constructive engagement, and then everyone inside it was blamed for behaving exactly as the structure incentivized them to behave.
This is particularly true of our often-rudimentary understanding of coalition politics. Boards, politics, marriages are all often “designed” by virtue of sheer habit, of experience, of no-one really thinking about the mechanics of a system. We do things because this is the way we always did it. Everyone knows that is the way this works, right?
This is the central claim of what I call conflict architecture: the discipline of designing conflict systems boards, teams, organizations, coalitions, even polities - so that the inevitable tensions of human collective life are channeled into productive channels rather than destructive ones. It is the difference between the firefighter, who arrives after the blaze, and the architect, who designs the building so the blaze never takes hold.
The distinction matters because our field has historically been oriented toward the back end of conflict. We are trained to mediate, to negotiate, to facilitate, to reconcile, to compromise, all of it after the rupture has occurred. That work is essential and will never be obsolete. But it is also, in a precise sense, too late.
We start playing catch-up, we give away things we would rather have wanted to keep. Resentment, distrust, conflict rigidity sets in. One conflict cause turns into three or four. Soon, it is difficult to get to the original challenge. By the time a mediator is called into a boardroom, the design flaws that produced the crisis have already done their damage. The most sophisticated intervention in the world cannot compensate for a structure that was built to fail, that fails to recognize its own design flaws. Conflict architecture inverts the sequence. It asks, before the first meeting is convened, before the first coalition agreement is signed, before the first team is assembled: what will this system do when it is under stress? And it designs accordingly, using the best in conflict systems design, an essential but underutilized new field in conflict studies and practice.
The anatomy of a design failure: the dysfunctional board
The board is the purest laboratory for studying conflict architecture, because a board is a system stripped to its essentials: a small group of individuals, bound by fiduciary duty, making consequential decisions under conditions of imperfect information and competing interests. When a board fails, the failure is almost never a mystery. It follows a recognisable pattern, and that pattern is structural.
Consider the classic dysfunctional board. It is not populated by bad people. It is populated by people who have been placed in a structure that guarantees friction without providing any mechanism for resolving it. The symptoms are familiar to anyone who has worked in governance: decisions that are endlessly deferred; information that is hoarded by an executive who treats the board as a rubber stamp; directors who disengage because they have learned their voice does not matter; a chair who conflates their role with the CEO's; factions that form around personalities rather than principles; and, beneath it all, a simmering resentment that occasionally erupts into open warfare. Each of these symptoms traces back to a conflict design flaw, not a character or motivational flaw.
Let’s look at a few of the more prevalent items of evidence:
Unclear decision rights:
The single most common source of board conflict is ambiguity about who decides what. When the boundary between governance and management is blurred, when the board micromanages operations one month and abdicates entirely the next, every decision becomes a contest for control. The conflict is not between people; it is between two people who both reasonably believe the decision is theirs.
Information asymmetry:
A board that receives its information exclusively from the executive it is supposed to oversee is structurally incapable of independent judgement. The executive controls the narrative, and the board, starved of alternative data, becomes either compliant or paranoid. The conflict that follows - the sudden, explosive challenge to a decision the board was never properly briefed on - is the predictable result of a design that concentrated information in one center.
The absence of structured dissent:
Healthy boards disagree. Dysfunctional boards either suppress disagreement until it becomes toxic, or allow it to run unstructured until it becomes personal. The design question is not whether there will be dissent but how it will be held. A board without a designed mechanism for dissent, such as a formal devil's advocate role, a culture of pre-mortems, a norm of separating the person from the position, some level of conflict training for senior executives and so on, will eventually experience dissent as betrayal, because that is the only form the structure permits.
Unmanaged status and identity dynamics:
Boards are not just decision-making bodies; they are arenas of status. Who speaks first, whose expertise is deferred to, whose identity is centred and whose is marginalised - these are design features, whether or not anyone intended them. A board that has never examined its own identity dynamics will find those dynamics surfacing as conflict, usually disguised as a dispute about something else entirely. The lesson of the dysfunctional board is that conflict is a symptom. It is the visible expression of a system that was never designed to carry the load it was asked to bear. Treat the symptom - mediate the dispute, replace the chair, send everyone on a team building retreat or conflict coaching program - and the underlying structure remains, ready to produce the next crisis on schedule.
The loads that a system or framework should be able to carry
If conflict architecture is the discipline of designing for stress, then the first task of the conflict architect is to understand the loads. What are the forces that every human system must carry, and where do they concentrate? I identify four, each of which maps onto a distinct body of theory and practice. This is, in practice, expanded upon and enhanced by Dispute System Design, a tool that should be expanded urgently in South African conflicts (see reference article below).
(i) The task load: organizational conflict
The foundational insight of organizational conflict research is that not all conflict is the same, and conflating its types is itself a design error. The distinction between task conflict (disagreement about what to do and how) and relationship conflict (interpersonal friction, animosity, distrust) is now well established. Task conflict, managed well, is a source of better decisions; relationship conflict is almost always corrosive.
The tragedy is that the two are not independent. Poorly managed task conflict becomes relationship conflict a disagreement about strategy curdles into a personal feud, and once that happens, the cognitive benefits of dissent are lost and the emotional costs remain.
The architect's job is to design systems that keep task conflict in the task domain. This is where psychological safety (Amy Edmondson's term for a climate in which people can speak up, admit error, and challenge authority without fear of punishment) becomes a structural requirement rather than a soft nicety. Psychological safety is not a feeling; it is a property of a system.
It is produced by specific design choices: how mistakes are handled, how dissent is received, whether the leader models fallibility, whether challenge is rewarded or punished. A team without psychological safety will convert every task disagreement into a relationship wound, because the structure gives people no safe way to disagree. Several modern, tested design solutions exist – conflict coaching, internal informal mediation and so on.
(ii) The team load: dynamics and interdependence
Teams carry a second load: the load of interdependence. The more a team's work is genuinely interdependent, or the more each member's success depends on the others, the more the team is exposed to the classic pathologies of group life and the near-inevitable group conflict dynamics, of which identity conflict is the operating system.
Underperformance, groupthink, unsustainable compromises and other poor conflict outcomes become predictable results where the desire for consensus overrides the capacity for critical evaluation. The Abilene paradox (as one example), where a group collectively agrees to a course of action that no individual member actually wants, because each misreads the others' silence as assent, becomes another predictable result and yet, if the design flaw is not spotted and repaired, it is destined to repeat itself. These are not failures of individual will. They are emergent properties of group structure.
The conflict architect's response is to design against them: to make roles and accountabilities explicit so that responsibility cannot diffuse; to build in structured dissent so that consensus cannot harden prematurely; to create norms and rituals that surface silent disagreement before it becomes a decision everyone regrets. As I regularly advise, we do not need less conflict, we need better conflict, and from this follows that the best teams are not the ones with the least conflict; they are the ones whose structure makes conflict productive.
(iii) The strategic load: complexity and paradox
At the highest level, systems carry the load of strategic complexity, or the condition in which the problems themselves are ill-defined, the values in tension are legitimate, and there is no single right answer. These are what the policy literature calls wicked problems: problems that cannot be solved, only managed, and where every intervention reshapes the problem itself. Strategic complexity produces a distinctive kind of conflict: conflict between goods, not between good and evil. The finance director who insists on fiscal discipline and the operations director who insists on investment are not enemies; they are both right, and the system must work with both approaches simultaneously.
This is the domain of conflict paradox: the recognition that organisations must pursue competing demands (exploitation and exploration, control and autonomy, short-term and long-term) that cannot be reconciled, only balanced. A system that is not designed for paradox will resolve it destructively: one side wins, the other is silenced, and the silenced side's legitimate concerns resurface later as sabotage, disengagement, underperformance or open revolt. A system designed for paradox creates a container for the tension - a forum, a process, a set of decision rules - that allows competing goods to be held in productive tension rather than resolved by domination.
(iv) The identity load: the deepest current
Beneath all of these runs the deepest load of all: the load of identity. And it is here that the conflict architect must be most precise, because this is the terrain where the discipline is most often misunderstood. Identity conflict is an immensely complex conflict arena, and I have written on this extensively, and readers who wish to know more about this essential modern conflict set of levers can have a look at the examples I have linked to below.
Identity conflict, not identity politics
There is a crucial distinction that is too often collapsed, and collapsing it does real damage to our ability to design for conflict. Identity conflict is a universal human phenomenon. Identity politics is a particular, instrumental strategy for mobilising it. The two are not the same, and a conflict architect who cannot tell them apart will design for the wrong thing. Identity conflict, in the sense developed by scholars from John Burton to Jay Rothman, from Jonathan Haidt to Jay van Bavel, arises from the fact that human beings have fundamental needs that are not negotiable: the need for recognition, for belonging, for dignity, for security of the self, for a coherent story about who one is and where one belongs.
These needs are not interests. Interests can be traded, compromised, bargained over. Needs cannot. When a conflict touches identity - when people feel that their very sense of self, their group's worth, their place in the world is at stake (often experienced as an existential threat), it behaves differently from a dispute over resources. It is more intense, more intractable, and more resistant to the rational, interest-based tools that dominate conventional conflict negotiation.
Rothman's ARIA framework (Antagonism, Resonance, Invention, Action) is built precisely on this insight for our study here. It begins from the recognition that identity conflicts are often mistaken for resource disputes, and that the first act of the practitioner is to correctly diagnose what kind of conflict is actually in the room. The Antagonism phase does not rush to suppress the conflict but seeks to understand it; the Resonance phase creates the conditions for parties to hear each other's needs and fears; the Invention phase opens space for co-creating possibilities that could not have been imagined from within the original positions; and the Action phase moves from insight to changed relationship.
DSD anticipates that this will happen, and provides a controlled, managed response and solution to it. The entire framework rests on a single premise: identity needs cannot be negotiated away, but they can be recognized, and recognition is itself transformative. Identity conflict does not respond effectively to conventional conflict strategies, and those seemingly good solutions can in fact be counterintuitively harmful.
This is the principle the conflict architect must internalize: every system carries identity, and every system that ignores identity will eventually be broken by it. A board that has never asked whose identity is centered and whose is marginalized, a team whose members feel their professional identity is threatened by a reorganization, a coalition whose constituent parties feel their distinct identities are being subsumed, or a nation whose citizens feel their group's dignity is not recognized by the state. In each case, the conflict that eventually erupts will be misdiagnosed as a dispute about resources or policy, when it is in fact a cry or demand for recognition.
To return to the popular misconception around the distinction from identity politics. Identity politics, as the term is used in contemporary debate, refers to the instrumental mobilization of identity - the deliberate activation of group identity as a political resource, the organization of people around race, ethnicity, religion, or culture for the purpose of gaining or holding power. It is a strategy, and like all strategies it can be used well or badly, for inclusion or for exclusion, for justice or for domination. The conflict architect is not in the business of endorsing or condemning identity politics.
The conflict architect's concern is more fundamental and more universal: identity is a load that every system carries, whether or not anyone is mobilizing it politically. The question is not whether identity will be present - t always is, by definition - but whether the system is designed to recognize and dignify identity, or to ignore and suppress it. A system that ignores identity does not make identity go away; it drives identity underground, where it festers and returns as conflict. We see this written all over our current political landscape. A system that dignifies identity, that gives every group a legitimate place, a voice, a sense of being seen, does not eliminate identity conflict, but it transforms it from a threat to the system's survival into a source of its richness, into my “better conflict” canvas. This is why the South African political case is so instructive, and so urgent.
The South African case study: politics in transition
South Africa is, at this moment, one of the most consequential live experiments in conflict architecture on the planet. The 2024 national election ended three decades of African National Congress majority rule and inaugurated a Government of National Unity (GNU), a ten-party coalition anchored by the ANC and the Democratic Alliance. It is a political system that has, in a single generation, moved from the most elaborately designed system of identity-based oppression in modern history (apartheid, which was nothing if not a triumph of destructive conflict architecture) to a constitutional order explicitly built around the rejection of that design, and now to a fragmented, coalition-based politics in which the old certainties no longer hold. This is about to be tested severely again in the next weeks and months, given the latest local election results. The stakes could hardly be higher, and the design questions could hardly be more pointed.
The coalition as a designed system: The GNU is, in structural terms, a board of directors for a nation, a group of parties with competing interests, distinct identities, and no single dominant authority, bound together by a negotiated agreement. The early evidence is precisely what conflict architecture would predict. The South African Institute of International Affairs has documented that while coalitions can enhance representation and accountability, their current operation is "marked by instability and opaque bargaining," with disputes between partners frequently spilling into the public arena and parties turning to the courts "amid a lack of transparent conflict resolution mechanisms." In other words: the coalition was assembled, but the system for managing its inevitable internal conflict was not designed (or everyone was blissfully unaware of such a possibility).
The parties signed an agreement about who would govern; they did not design a mechanism for what would happen when they disagreed, or they put in place a rudimentary, outdated perceived solution to the inevitable intractable conflict on the horizon, which is to say, they did not design for the one thing that was certain to occur. This is the dysfunctional board, scaled to national or citywide proportions.
The symptoms are identical: unclear decision rights (who has the final say on contested policy), information asymmetry (which partners are briefed, and when), the absence of structured dissent (disagreements conducted through public statements and litigation rather than through a designed forum), and unmanaged identity dynamics (the ANC's liberation-movement identity, the DA's liberal identity, the IFP's Zulu-nationalist heritage, the MK party's mobilisation of KwaZulu-Natal grievance etc. with each a distinct identity load that the coalition structure must carry).
The identity load of a nation: South Africa's identity load is heavier than almost any other democracy's, because the country was designed around identity. Apartheid was a system that assigned every person a racial category and built the entire social order on those categories. The post-1994 constitutional order was, in part, a deliberate act of counter-architecture: a design intended to create a non-racial society, to give every identity a legitimate place, to recognise the dignity of all. The 2025 Reconciliation Barometer from the Institute for Justice and Reconciliation finds that national identity remains strong - a testament to the resilience of that design - even as it documents the strains: political tension and violence in KwaZulu-Natal, intense partisan contestation, the Afrikaner wars, and warnings of instability ahead of the 2026/27 local government elections.
The contemporary challenge is that identity is being re-mobilized. Analysts and jurists have warned that race, ethnicity, and tribal affiliation are again becoming powerful political markers, and that this trend "could undermine the constitutional vision of a non-racial society." This is the precise point where the identity-conflict/identity politics distinction becomes operationally vital. The conflict architect does not respond to the re-mobilisation of identity by denying that identity matters, as that would be to repeat the error of a system that ignores identity and drives it underground, or as we see so often in public, is not understood at any meaningful depth by decision makers. Nor does the conflict architect simply capitulate to identity politics.
The question is structural: is the system designed to effectively manage identity, or to be captured by it? A political system that truly understands and effectively manages identity gives every group a legitimate, recognised place, without the harmful compromises politics we are seeing, in its institutions, its symbols, its distribution of power and resources, while refusing to let any single identity become the basis for excluding others.
An organization or coalition captured by identity politics becomes a zero-sum contest in which each group's gain is another's loss, and the constitutional vision of a shared, non-racial citizenship is progressively hollowed out, or it becomes a wasteland of increasingly harmful and resentful compromises. The difference between these two outcomes is not a matter of rhetoric or goodwill. It is a matter of design: of electoral systems, of coalition agreements, of the structure of local government, of who is included in the institutions of the state and how their inclusion is made real.
The local government fault line: The most acute design failure in contemporary South Africa is not at the national level but at the local. The dysfunction of the local government sphere, documented repeatedly and acknowledged across the political spectrum, is a textbook case of conflict architecture gone wrong. Municipalities are where the identity load, the resource load, and the governance load all concentrate at once: where service delivery failures meet identity-based grievance, where coalition instability meets community frustration, where the abstract tensions of national politics become the concrete, lived experience of citizens.
The 2026/27 local elections, with nearly half of South Africans telling Ipsos that no party represents their views, will be a stress test of the entire system. A political system that has not designed its local government to carry these loads will see them expressed as protest, as violence, as the slow erosion of the very legitimacy on which the constitutional order depends.
The Conflict Architect's Code: five design commitments
What, then, does the conflict architect actually do? The discipline can be distilled into five design commitments, five questions that must be asked of any system before it is built, and five properties that must be designed into it if it is to carry its loads without breaking. The hard yards come from Dispute Systems Design theory and practice, but as a start we can consider the following five guiding commitments:
1. Design for load
Anticipate where stress will concentrate, and build for it. Every system has predictable stress points: the boundary between governance and management, the allocation of scarce resources, the moment of leadership transition, the clash of competing identities. The architect maps these in advance and designs the structure to carry them, rather than discovering them in the form of a crisis. A coalition agreement that does not specify what happens when partners disagree is a structure built without load-bearing walls. Coalitions will test this very soon, and in the absence of “coalition legislation” these solutions are the perfect solutions for the tasks ahead.
2. Design for voice
Ensure that every stakeholder has a legitimate, structured channel for dissent. Conflict becomes destructive when voice is blocked, because blocked voice does not disappear - it goes underground and returns as sabotage, disengagement, or explosion. The conflict architect builds in the mechanisms of voice: clear decision rights, formal dissent roles, psychological safety, forums where disagreement can be held without becoming personal, face saving mechanisms, identity conflict resolution mechanisms. The measure of a healthy system is not the absence of conflict but the presence of effective channels through which conflict can flow.
3. Design for dignity
Recognize that every system carries identity, and design for it explicitly. This means asking, of every structure: whose identity is centred, whose is marginalised, and is there a legitimate place for all? It means building recognition into the architecture - in representation, in symbols, in the distribution of voice and power - so that identity needs are met by the system rather than expressed against it. Dignity is not a soft value; it is an essential property. A system that denies dignity will be broken by the very people whose dignity it denies. It has nothing to do with political correctness and everything to do with sustainable efficiency.
4. Design for signal
Build early-warning feedback loops that surface conflict while it is still small, a sort of early warning system, if you will. Destructive conflict is almost always preceded by signals: rising grievance, declining trust, the quiet withdrawal of engagement, spite, public spats, that a well-designed system would have detected. The architect builds in the sensors: regular, honest assessment of the system's health; mechanisms for surfacing silent dissent; the willingness to treat an increase in reported conflict not as a failure but as a sign that the system is working, because conflict is being surfaced rather than suppressed. Again, dispute system design makes this point precisely: a rise in disputes after a new system is installed may indicate recovery, not failure, because it means conflict is finally being addressed constructively.
5. Design for repair
Build restorative pathways before rupture occurs. No system, however well designed, will avoid all conflict. The question is what happens after the rupture, whether there is a designed path back to relationship, or only the slow accumulation of unresolved wounds. The architect builds in the mechanisms of repair: mediation, reconciliation, the rituals and processes by which a system acknowledges harm and restores trust. A system without a designed path to repair is a system in which every conflict is permanent.
Conclusion
As indicated, the conflict architect does not believe that conflict can be eliminated, or that a perfect design will make human beings stop being human. What would we do without conflict? The conflict architect knows that tension is not a defect of collective life but its condition, that the very differences that make a board wise, a team creative, a coalition representative, a nation rich, are the same differences that make conflict inevitable. The conflict architect's ambition is therefore not to end conflict but to design for it, to build systems that are strong enough to carry their loads, open enough to hear every voice, dignified enough to recognize every identity, sensitive enough to detect the first tremor, and wise enough to repair what breaks.
This is the work that South Africa is now urgently called to do. Effective coalitions do not require legislation, trust, politeness or friendship. It has available to it a range of already tested conflict tools that work - accountability mechanisms, contingency agreements - a whole spectrum of solutions that accept that things may fall apart. It is the work of a generation. And it begins, as all architecture begins, not with the building but with the design, with the refusal to accept that conflict is something that merely happens to us, and the insistence that it is something we can, and must, design for.
Sources and further reading
Our blog at www.conflict-conversations.co.za has a rich variety of articles expanding on specific aspects of the article, and conflict in general.
Andre Vlok can be contacted at andre@conflict1.co.za for any further information.
(c) Andre Vlok
November 2026