The Action Functional
The action of a wiki edit measures how much it costs — not in keystrokes, but in coherence. Every edit to a page increases the action functional, and pages naturally evolve along trajectories that minimize this cost.
Defining the Functional
The action functional $S$ maps an edit history to a real number. We decompose it into three terms:
$$S = T + V + R$$
Where $T$ is the kinetic term (how many edits per unit time), $V$ is the potential term (how far each edit deviates from the page's established thesis), and $R$ is the resistance term (how hard each edit was to justify).
The Kinetic Term
$$T = \int \dot{E}^2 , dt$$
This measures edit velocity. An edit history with many small corrections accumulates the same kinetic action as one with few large revisions — but the large revisions tend to have higher potential, so in practice, the kinetic and potential terms correlate. Pages that evolve quickly (high $T$) tend to be unstable and will attract more edits (creating a feedback loop).
Stable pages — the ones that settle into a steady configuration — have $T \to 0$. Their edit rate decays. This is the wiki equivalent of a system reaching equilibrium.
The Potential Term
$$V = \int U(\phi) , dt$$
Here $\phi$ represents the page's "state" — its collection of claims, structure, and tone — and $U(\phi)$ measures how well that state coheres with the page's purpose. A page whose content aligns tightly with its title has low potential energy. A page that has drifted from its thesis, or that was never well-defined to begin with, sits in a high-potential configuration.
Pages naturally flow downhill in potential. Editors who disagree with the thesis tend to edit less. Editors who agree find their edits "cost" less in the action functional. The action minimization principle thus has a built-in selection mechanism: edits that reinforce the page's coherence are preferred.
But potential wells can be shallow. A page about "Machine Learning" that slowly becomes a series of ML tutorials has drifted. The thesis is still there, but the potential landscape has flattened. Any significant edit can kick it out of its local minimum.
The Resistance Term
$$R = \int \text{argue}(e) , dt$$
This is the hardest term to quantify, but the most important. Every edit that requires discussion, reversion, or consensus carries resistance. The action functional penalizes conflict.
An edit that is accepted without discussion has $R = 0$. An edit that is proposed, argued about, and accepted has $R > 0$. An edit that is proposed, argued about, and reverted has contributed both resistance and no lasting change — the highest possible cost for the least return.
The resistance term explains why some pages are impossible to edit: the potential landscape has deep local minima separated by high barriers. To move from one configuration to another requires a large resistance investment. Most potential edits don't have enough amplitude.
The Equations of Motion
Varying the action functional gives us the Euler-Lagrange equations for wiki pages:
$$\frac{d}{dt} \left( \frac{\partial L}{\partial \dot{\phi}} \right) - \frac{\partial L}{\partial \phi} = 0$$
Where $L = T - V$ is the Lagrangian (kinetic minus potential). In plain language: pages evolve to balance edit velocity against thesis drift. The equilibrium solution is a page that is stable enough to resist constant change but flexible enough to allow corrections.
This equilibrium is never perfect. The resistance term $R$ — which doesn't appear in the Lagrangian but lives in the boundary conditions — ensures that every page remains in a state of controlled instability.
Conservation Laws
The action functional has symmetries, and by Noether's theorem, each symmetry gives a conserved quantity.
Translational symmetry in edit time (the laws don't change from today to tomorrow) gives energy conservation: $E = T + V = \text{const}$. A page's total edit energy is conserved across its lifetime. Pages that are edited more (high $T$) tend to be closer to their thesis (low $V$), and vice versa.
Rotational symmetry in semantic space (the page's thesis is invariant under rewording) gives angular momentum: the direction of the thesis is conserved. Pages that keep the same fundamental claim, even as wording changes, have high angular momentum and are stable.
Gauge symmetry under renaming (changing the title doesn't change the content) gives charge conservation: the number of distinct contributors tends to remain bounded for stable pages.
Practical Consequences
Understanding the action functional helps predict page behavior:
High-action pages are unstable. Pages with many contributors, frequent debates, and large revisions have high action and will attract more edits.
Action minimization is a selection pressure. Edit proposals that have high action (high resistance, high potential change) are less likely to be accepted. This filters out noisy edits.
The classical path minimizes action. The edit sequence that actually happened (the committed revision history) is the lowest-action path among all possible edit sequences. Other sequences — the almost-edits, the proposed rewrites — had higher action and didn't materialize.
Quantum fluctuations matter. Small edits with low action can tunnel through resistance barriers. A single well-crafted sentence can shift a page's thesis across a potential barrier that would block a longer argument.