In our previous article, we re-imagined hierarchies not as structures of domination or authority, but as structures of access: cumulative, threshold-based systems that expand the set of options. Each level of the hierarchy unlocked new possibilities without revoking the old, forming a scaffold for growth, capability, and decision-making.
But this raises a deeper question:
What do hierarchies do once they are adopted? How do they shape thought, action, and behavior in practice?
In this article, we explore that question, and in doing so, we uncover a more general principle:
A hierarchy, once adopted, functions as a meta-strategy.
That is, it operates not as a direct strategy itself, but as a higher-order constraint: a filter over the space of all possible strategies, selecting only those that conform to the structure of access defined by the hierarchy.
This leads us to the broader topic at the heart of this article: the concept of meta-strategy.
In the sections that follow, we define what a meta-strategy is, showing how a hierarchy is simply a type of meta-strategy.
Let $\Pi$ represent the strategy space: the full set of possible strategies an agent might follow in a given domain.
A meta-strategy $M$ is a higher-order constraint or function acting on $\Pi$, producing a subset $\Pi_M \subseteq \Pi$ that represents the valid or preferred strategies under that meta-constraint:
\[M: \Pi \to \mathcal{P}(\Pi), \quad \text{with} \quad \Pi_M = M(\Pi)\]Here, $\mathcal{P}(\Pi)$ denotes the power set of $\Pi$: the set of all possible subsets of the strategy space. This means a meta-strategy does not output a single strategy — it selects a subset of strategies that are permissible, plausible, or desirable under its constraints.
In simple terms, a meta-strategy does not directly choose an action — it shapes the space in which choices can occur. It filters.
This concept shows up across domains:
One clear and structured example of a meta-strategy is a hierarchy.
As described in our earlier work, a hierarchy is not a ladder of power, but a structure of access: a series of thresholds, each expanding the available set of options.
When adopted, a hierarchy $H$ acts as a meta-strategy $M_H$, filtering the global strategy space $\Pi$ down to only those strategies that respect its layered, cumulative constraints:
\[\Pi_H = M_H(\Pi) = \{ \pi \in \Pi \mid \pi \text{ conforms to } H \}\]That is:
When applied, a hierarchy acts as a meta-strategy.
This mirrors design patterns in software architecture: an abstract class defines a structural framework without dictating implementation — it tells you what must be there, not how it must work.
In the same way, a hierarchy shapes strategy without specifying every detail. It is not a script, but a scaffold.
The central insight is this: meta-strategies constrain the space of possible strategies. Rather than specifying particular actions, they define the conditions under which strategies are considered valid or permissible.
Hierarchies exemplify this principle. When adopted, a hierarchy does not prescribe actions directly. Instead, it functions as a meta-strategy — a higher-order constraint that filters the global strategy space $\Pi$ to a subset $\Pi_H$ that aligns with its structure of access.
This perspective shifts our understanding of hierarchies — and meta-strategies more broadly — from static frameworks to active filters that shape decision-making. A meta-strategy structures how strategy selection occurs, thus influencing not only outcomes but the very space of possible moves.
Understanding this helps clarify how systems, policies, heuristics, and even values operate at a higher level: by governing the strategic landscape itself.