The tools that generic mediation does not give you
Generic conflict resolution will fail you in a mining dispute. The reason is structural, not personal. A standard mediation training prepares you to manage two parties with roughly comparable standing who want a faster settlement than a court can give. An extractive conflict gives you none of that. You face a multinational operator, a community that did not choose the project, and a regulator pulled between revenue and constituents. The parties hold unequal money, unequal expertise, and unequal time. The agreement you broker must survive years of construction and operation, not a single signing.
That is why you need a defined set of tools built for this context. You need ways to read power across more than one dimension. You need a structure for splitting a tangled dispute into parts that each behave differently. You need methods for helping a community speak with one legitimate voice. You need agreements specific enough to monitor. This article walks through the working tools, grouped by where they fit in a case. The downloadable selector at the end pairs each technique with the moment it earns its place.
Diagnostic tools: read the field before you convene
Every extractive mediation starts with a reading of power, and most readings are too shallow. Stated positions tell you what parties say they want. They do not tell you who can actually move. Treat power as four separate things. There is positional power, meaning authority and control over resources and permits. There is relational power, meaning trust and networks. There is knowledge power, meaning technical command and control of information. There is cultural power, meaning legitimacy and the right to speak.
Map each party across all four. A company often holds strong positional and knowledge power while lacking cultural legitimacy in the affected area. A community often holds cultural and relational power while lacking technical command of mining hydrology or air-quality science. A regulator may hold formal authority while being squeezed by competing pressures. Once you see the gaps, you see your job. Where a community cannot evaluate a water claim, you broker access to an independent expert. Where a company cannot read the local social map, you design engagement that builds standing it does not yet have.
The second diagnostic is governance mapping. Communities are not single parties, and treating them as one is the most common failure in this work. A project touches landholders, farmers, traders, wage laborers, women, youth, and elders, each with distinct stakes. Before you convene anyone, document how the community actually reaches collective decisions. Identify who holds customary authority and who speaks for each sub-group. Formal recognition of a leader does not mean members accept that leader’s authority. Get this wrong and any agreement you reach will not hold, because the people bound by it never authorized it.
Process tools: design a table that fits the dispute
Extractive disputes bundle several problems into one. A fight over water quality is rarely about water alone. It pulls in agriculture, livestock, health, the funding of monitoring, the compensation for unavoidable impact, and the rule for adjusting if conditions change. Negotiating all of that at one table guarantees deadlock. Segment instead. Break the dispute into technical, financial, and relational components, and handle each with the structure it needs.
Technical components are disagreements about facts. The company says mining will not lower the water table. The community says it will. You do not bargain a fact. You commission independent investigation, agree in advance how the finding will bind the talks, and remove the question from the relational arena. Financial components behave differently. Compensation, benefit sharing, and remediation funding distribute a finite pool, so traditional bargaining applies. You help the parties size the pool, surface the claims on it, and fix a formula and a payment mechanism. Relational components are the longest-lived. They cover how decisions get made jointly, how concerns get raised, and how the agreement adapts over time. Here your job is institutional design, not a one-time deal.
Reality testing belongs in your process kit, because parties often arrive holding expectations no agreement could meet. A short, respectful examination of a party’s alternatives keeps the table honest, and it works far better than argument. I treat it as a craft skill rather than a confrontation, the same way I treat the reading of facts against emotions in a heated room. The goal is not to win a point. The goal is to let a party see for itself what walking away actually costs.
Process tools: support the community’s own negotiation
Your role is not to negotiate with the community. Your role is to help the community negotiate with the company. That distinction changes everything you do. It means you protect the community’s internal process rather than substitute for it. Four elements carry this work.
First, representational support. Help the community set up clear, transparent channels for representatives to consult and report back. Designate points of contact, schedule consultation meetings between sessions, and confirm that representatives carry real authority to negotiate. Second, independent advice access. The community should reach independent technical, legal, and financial counsel to assess what the company proposes. You help identify advisors and arrange the logistics for them to take part. Third, deliberation time. Build pauses into the schedule so the community can consult internally after each major session. This is not delay. This is the work that produces a legitimate collective decision.
The fourth element is your own discipline about pace. A community that feels rushed will either reject a fair deal or sign one it cannot defend at home. Both outcomes fail. Slowing down to let internal deliberation finish is one of the most effective moves you have, and it costs nothing but patience.
Communication tools: carry the parties through the hard moments
The room will get difficult, and you need techniques for those moments. Separate sessions let you hear what a party will not say in front of the other side, and they let you test a concession before it is exposed. Reframing turns a trust contest back into an answerable question. When two sides are really fighting about whether they can believe each other, you name the factual question underneath and route it to an expert. The dispute becomes solvable again.
Acknowledgment sequencing matters more than newcomers expect. In many cases, an unaddressed grievance or a felt disrespect blocks every move on substance. Until someone names the harm and the other side hears it, no number on a page will land. This is not about admitting legal liability. It is about restoring enough standing for a conversation to proceed. The craft of holding that moment is part of what separates a capable mediator from a trained one. Reading when the room is ready, and when it is not, is exactly the judgment I develop in the seven core qualities that distinguish exceptional mediators.
When direct contact between parties does more harm than good, you move information rather than people. Shuttling proposals between rooms keeps a process alive that a single joint table would have killed. I use it deliberately, with clear rules about what I carry and what I hold, as set out in this guide to shuttle diplomacy in mining conflicts. Used well, it lowers the temperature without lowering the stakes.
Agreement tools: make the deal hold for years
Most mining agreements succeed or fail in implementation, not in negotiation. A strong deal collapses when commitments are too vague to monitor or when conditions shift in ways nobody planned for. So your last set of tools protects the outcome. Start with an implementation specificity check before anyone signs. A clause that says the company “will conduct water monitoring” is not implementable. A clause that names the parameters, the testing frequency, the standards, the tester, the reporting path, and the triggers for action can actually be enforced. Drive every commitment to that level of detail.
Pair this with a remedy and causation worksheet whenever harm is claimed and disputed. Separate the evidence question from the design question. First document the specific harm. Then establish how much is attributable to the operation rather than to drought, traffic, or other sources. Then size the severity. Only then design the remedy. Keeping causation distinct from remedy is what unsticks talks that would otherwise circle a blame argument forever.
Then build the standing structure. A joint monitoring committee, with company, community, and an independent member, meets on a fixed schedule to review progress, log concerns, and settle small disputes before they grow. Give it authority over minor issues and a clear referral path for major ones. Add a built-in dispute resolution clause so future disagreements follow an agreed route rather than reopening the whole settlement. These structures move a relationship from adversarial to durable, which is the shift I describe in turning mining conflict dynamics from adversarial to collaborative.
Seeing the toolkit work together
Consider a scenario drawn from patterns across West African gold projects. An operator begins work near farming villages and artisanal mining sites. Within months, disputes erupt over water access, displacement, and dust. The company sees obstruction. The communities see disregard. The talks are stuck.
A mediator applies the kit in sequence. The power map shows the company strong on permits and technical command, weak on legitimacy, and the communities strong on standing, weak on technical command. So the mediator brokers independent water and dust assessments, funded by the company but run by consultants the communities choose. That closes the knowledge gap and gives everyone a shared factual base. Governance mapping produces legitimate representatives from the farming, artisanal, and general groups, each consulting constituents between sessions. Issue segmentation splits the dispute into three strands. A technical strand runs on the assessments, a financial strand covers compensation and livelihood support, and a relational strand becomes a joint monitoring committee. The remedy worksheet shows water impact mostly attributable to mining and dust attributable to both mining and local traffic, which lets the parties assign responsibility precisely. The resulting agreement is specific, monitored, and, in this illustrative pattern, holds for years with only minor issues reaching the committee.
Match the tool to the moment instead of running them all at once
The Extractive Mediation Toolkit Selector, a companion tool, is a two-column reference that pairs each technique above with the moment that should trigger it. Read each row as a technique plus its trigger, and go straight to the group that matches where your dispute sits: “Diagnostic Tools (Before You Convene)” for power mapping and governance mapping, then the process-stage group for issue segmentation, reality testing, and shuttling, and the agreement-stage group for the specificity check, the remedy and causation worksheet, and the joint monitoring committee. When several triggers apply at once, which is the usual case in a live file, the selector tells you to sequence the techniques rather than deploy them together. It keeps you from reaching for a table redesign when the real block is a factual dispute you should have sent to an expert. Download the Extractive Mediation Toolkit Selector.
The case for mediation, and a structure to deliver it
These tools share one premise. An extractive dispute settles best when a skilled, independent third party facilitates it, not when the parties litigate, escalate, or wait it out. The adversarial default is slow and expensive, and it tends to harden the very relationship a project depends on for the next twenty years. Franks et al. (2014) documented how company-community conflict converts environmental and social risk into real business cost, much of it through delay and lost time. Mediation attacks that cost directly. It keeps the parties talking, narrows the dispute to answerable questions, and builds the standing structures that prevent the next conflict.
A toolkit alone is not a method. The tools need an architecture that sequences them and ties each technique to a durable agreement. That is the work of the Social Accord Architecture, my framework for converting fragile settlements into agreements that survive the life of a project. The Social Accord Architecture organizes diagnosis, process, and agreement into a single discipline, so a mediator is not improvising tool by tool but running a coherent system. The Trust Audit reads where standing has broken. The Blueprint specifies the agreement so it can be monitored. The Resilience Handover passes a working relationship to the people who must live with it. If you mediate in this sector, build your practice on a structure like the SAA, and reach for the right tool when its moment arrives. To talk through how these tools fit a specific dispute, write to thomas@thomasgaultier.com.
