A familiar scene: the board is deep into a quarterly meeting when a director leans forward and says, "I'm not comfortable approving this investment until we've reviewed the underlying implementation plan in detail." Another director, equally experienced, pushes back: "If we're getting into the weeds, we're not governing. We're managing."
The chair tries to land the moment with the old line: "noses in, fingers out."
That phrase gets repeated in boardrooms because it points at a persistent tension. But it doesn't tell you how to manage it. When is "noses in" wise oversight, and when is it theatre? When is "fingers out" healthy restraint, and when is it abdication of responsibility?
In this sixth and final piece I explore how polarity mapping earns its place at the board level. Not by turning governance into a workshop, but by giving directors a disciplined way to do what boards are structurally designed to do: hold interdependent opposites in productive tension over time.
To reiterate the central thesis of this series: some tensions are not problems to solve (choose A or B), but polarities to manage (get the upsides of both, avoid the downsides of either). These are insoluble problems that require ongoing management rather than one-off decisions.
Why Boards Are Polarity Machines (whether They Realise it or Not)
If you strip corporate governance back to its intellectual roots, you can see why polarities are central.
Agency theory starts from a sober premise: the separation of ownership and control creates incentives for managers to pursue their own interests, so governance mechanisms—boards included—exist to monitor and align these self-interests with the best interests of the company. This logic remains deeply influential in law, investor expectations, and governance codes. The G20/OECD Principles, for example, explicitly pair strategic guidance with effective monitoring as core board responsibilities.
The corporate governance framework should ensure the strategic guidance of the company, the effective monitoring of management by the board, and the board's accountability to the company and the shareholders.
— G20/OECD Principles 2015
But boards are not only monitors. They also connect firms to external resources: legitimacy, networks, information, and influence. A reality which can most succinctly be called board capital.
Even the behavioural view of boards points in the same direction with boards often described as "large, elite, and episodic decision-making groups" dealing with complex strategic-issue processing—meaning board effectiveness depends heavily on how directors think and interact.
This casts the board's job as inherently double-sided—or perhaps it's better to say it is double-edged—and requires directors balance notions of independence and involvement; challenge and support; control and collaboration. Governance, in short, isn't a stable position. It's a form of ongoing navigation.
That is precisely the territory of polarity mapping.
The Board's Master Polarity: Control ↔ Collaboration
If you want one polarity that shows up in almost every board dysfunction story, it's this:
Control (monitoring, discipline, accountability) ↔ Collaboration (support, advice, empowerment)
Corporate governance has been described as a clash between control and collaborative approaches, and in consequence is better understood as a series of reinforcing cycles that can produce decline if overemphasised but which cannot be ignored as they are needed for long-term vitality. This framing is particularly useful for directors because it also bridges agency and stewardship logics rather than treating them as mutually exclusive.
If we take this tension further by modelling the board's dual role as advisor and monitor we see an uncomfortable core insight: when boards monitor more intensively, CEOs may be less willing to share information—reducing the board's ability to provide good advice. In other words, "more oversight" can paradoxically produce less oversight, because it starves the board of candour.
Here's what the board-level polarity map looks like when made explicit.
