Universality is the simplifying move in strategy

- strategy isn't complication but our discourse has become

There are two objections that come, like clockwork, whenever I press the case for a universal definition of strategy. The first is the proud sceptic’s refrain: strategy is too knotted, too context-corded, too situation-stained to be held within a single theory. The second is its practical cousin: grant that some universal definition might exist, yet QCEA and SO(X) framing is too baroque to be of use.

I want to turn these objections on their heads. The universal hypothesis is not the complicated one. It is the simplifying one. What has become complicated is our business discourse, because we have allowed product narratives to masquerade as definitions. We take a pleasing sound bite like, “Strategy is deciding where to play and how to win,” and we treat it as if it were a definition of the phenomenon. Then, a little later, we tighten it again into something like, “an integrated set of choices that compels desired customer action.” Both are rhetorically neat. Both are commercially useful. Both point toward real aspects of competitive life. And yet notice what has happened: we have not converged on a stable definition of what strategy is. We have iterated the pitch.

This is not a moral moan; it is a structural note. A phenomenon that stands apart from our telling of it does not change merely because our fashions change. Predation did not shift when we christened it “pursuit–evasion”. Immunity did not quiver when we found fresh metaphors for it. Only our accounts altered. If “strategy” is real in the same manner, it should leave a stable footprint in the world. What changes is our rede of that footprint, and, more pointedly, the wares we vend beneath the banner of “strategy”: plans, portfolios, operating models, transformations, OKRs, culture programmes, analytics initiatives. Each is deliverable; each is packable; each can be propped up by a long chain of reasons. Lay those chains side by side and you begin to mistake the chains for the thing, and to conclude that the underlying phenomenon must be impossibly complex and irreducibly context-bound.

My wager is therefore sharper than parsimony. It is diagnostic. If we stop treating strategy as the artefact, and start treating it as the underlying function, the concept becomes smaller, not larger. QCEA and SO(X) are not complexity theatre. They are an attempt to compress what recurs, across all domains, into a minimal structure that can survive contact with reality.

But there is a trap waiting for us, and it is the trap that makes universal claims sound foolish: if everything is strategy, then nothing is. So before we look outward, we need ground rules. We must say what counts as strategy, and what does not.

Ground rules

Ere we wend anywhere near animals, markets, wars, or boardrooms, we need a boundary that can fail. Otherwise we slide into the slack comfort of calling everything “strategy”, and then congratulating ourselves for discovering it everywhere.

Here is the minimal rule set I use. Strategy is not a document, not a meeting, not a plan, and certainly not a genre of PowerPoint. Strategy is coherence-preserving inference-and-action: a state-contingent policy that selects among real alternatives under informational and resource constraints, with observable consequences. When the state changes, the policy can change. That capacity to switch is the signature. If the world becomes riskier, or resources tighten, or an opponent adapts, or a partner proves unreliable, and the system systematically changes what it does, that is the footprint of strategy. You do not need language for this. You do not need consciousness in the grand, theatrical sense. You need contingency.

Now the guardrail that saves us from the “everything is strategy” trap. Strategy requires real alternatives. There must be more than one viable policy available. There must be trade-offs under constraints. And there must be consequences you can observe, not just stories you can tell. If an organism, a firm, or a colony behaves in the same way regardless of cost, context, or threat, that is not strategy by this definition. It might be a reflex. It might be a fixed routine. It might even be an elegant routine. But it is invariant. It does not select among options, because it does not have options.

This matters because it makes the claim testable. Under this boundary, a great deal of behaviour is disqualified. Much of life is habit, pattern, and automaticity. Strategy is the special case where behaviour is conditional on state in a way that predicts outcomes. It is not “being adaptive” in the vague sense. It is not merely “trying”. It is not a synonym for ambition.

So the question becomes concrete. If we look across the broad empirical record, do we actually see this signature: alternative policies, constraint trade-offs, and state-dependent switching with observable consequences?

To keep this from becoming a zoo tour, we need a map. Not species by species, but problem-type by problem-type. That is where we wend next.

The map

With the ground rules in place, we can stop wrangling about corporate artefacts and start looking for the phenomenon itself. But we need to do it properly. If I simply rattle off species, we learn nothing except that the natural world is busy. The universality claim does not live or die on whether a raven looks “clever” or a wolf looks “strategic”. It lives or dies on whether the same problem-classes reliably elicit the same functional signature: alternative policies, constraint trade-offs, and state-dependent selection with consequences.

So here is the map. Not a taxonomy of animals, but a taxonomy of strategic burdens, the kinds of situations that force a system to choose among policies because constraints make “always do X” untenable.

First, there is adversarial interaction: pursuit and evasion, raids and defence, contests where the other side adapts and your success depends on tempo, coordination, and sometimes interference with what the other side can perceive. This is the oldest theatre of strategy, and it is the one people instinctively accept, because it smells like war.

Second, there is governance and coalitionary politics: who aligns with whom, who polices whom, how conflict is suppressed before it becomes ruinous, and how power is exercised without constant violence. You might dislike the word politics, but you recognise the problem. Any social system that endures must manage it.

Third, there is cooperation and exchange: the problem of mutual benefit under the perpetual risk of cheating. This includes partner choice, sanctions, and the quiet arithmetic of reciprocity. There is nothing sentimental here. Cooperation is often a hard-edged technology for stability.

Fourth, there is information strategy: signalling and concealment, credibility and deception, reputation and audience effects, even interference. It is the domain where what matters is not only what you do, but what others infer you will do next.

Fifth, there is logistics and buffering: the time-binding side of strategy. Stockpiling, redundancy, pre-positioning, and the protection of stores. When uncertainty is real, the future is not a date. It is a constraint, perhaps a shock, waiting to happen.

Sixth, there is mobility and search: navigation, routing, exploration versus exploitation, quorum decisions, and collective movement rules. This is strategy as the management of attention and effort across a landscape whose rewards are uneven and often hidden.

Seventh, there is capability extension and niche construction: tool use, environmental engineering, and the creation of new affordances. This is where strategy stops being merely the selection of moves within a given game and becomes the shaping of the game itself.

That is the scaffold. In each category, the question will be the same, and it will stay stubbornly unromantic. When constraints tighten or states shift, do we observe policy selection among real alternatives, with measurable consequences? If we do, the universality claim gets stronger. If we do not, it ought to fail, cleanly and without consideration.

Now let’s walk the map. Not metaphor. Mechanism. We begin where most people feel safest: adversarial strategy, in the raw world of pursuit, evasion, and interference.

A quick note to the reader before we do: Remember resolution. When we examine the empirical record across species, the phenomenon we are tracking, strategy, appears at different resolutions. A primate such as Homo sapiens is a strategic organism and can pursue strategy at the resolution of the individual. Yet that same primate is also a constituent actor within a human herd, what it likes to call an “organisation”, which may itself qualify as a strategic organism. The same is true for tamer species such as the lion, which can be strategic as an individual and also as a constituent actor within a larger strategic organism known, by the collective noun, as a “pride”.If your resolution is locked, your vision will blur quickly as you read. See earlier articles, written by the same pen-wielding primate who penned this mischievous missive, for a fuller account of resolution and the strategic organism.

Adversarial strategy without language

Start at night, because the tenebrous hours are honest. There are no slogans in the dark, no committees, no consoling narratives about “vision”. There is only a narrow margin betwixt feeding and failing. A Mexican free-tailed bat does not “know” the world in the way we flatter ourselves that we do. It samples it. It casts a pulse into the air and hearkens for the world’s reply. Call and echo, echo and call, a brisk little inference engine stitched to muscle and wing.

Now place another bat in the same slice of sky, hunting the same insects, and the problem changes shape. It is no longer only “find prey”. It is “find prey while someone else is also listening, also converging, also competing”. In that competitive state, some bats deploy a specialised call that disrupts a rival’s echolocation during prey pursuit. Sensor denial is the plain name for it: one system degrades another system’s sensing, and does so at the moment it matters. No boardroom is required for this to be strategic. The signature is enough: there are alternatives, there is a trade-off (spend energy on interference or spend it on pursuit), and there is state-dependent selection with observable consequences in the competitor’s performance.

Come up into daylight and trade the night air for the bright, briny glare of the thalassic world. A bait ball of sardines is a moving fortress, a silvered geometry of fear. Against a lone attacker it is hard to crack. The constraints are not philosophical. Time is short. Energy is dear. A mistake can mean injury. Yet in this theatre sailfish hunting in groups often alternate attacks on the schooling prey. Those repeated attacks injure fish frequently, and the accumulation of injury increases capture efficiency later. That is a form of tempo management, not as metaphor but as mechanism. A simple policy becomes visible: do not all strike at once and squander your advantage in chaos; strike in sequence, let the prey state deteriorate, and let each attacker benefit from the changing conditions created by the others. Again, you do not need to imagine a shared plan. You only need to see a contingent pattern that solves a constraint problem.

Now take it inland, into the brumal arithmetic of winter and hooves. Wolves hunting elk and wolves hunting bison are not solving the same problem with a different costume. They are solving different constraint regimes. The pay-offs and dangers change with prey type, and the marginal value of additional hunters changes with group size. The evidence here is not a romantic vignette of “teamwork”. It is more austere, and more useful: attack, selection, and kill success shift nonlinearly with pack size, and the scaling differs with the formidability of the target. In other words, the effective policy is not “hunt” but “hunt like this, with this many, against this quarry, at this risk”. That is strategy in its leanest form: state-dependent selection among alternatives under constraint, with measurable consequences.

Move again, back to water, but this time to ice. A seal on an ice floe looks like a problem of simple reach. It is not. It is a problem of medium, geometry, and timing. Some killer whale pods generate waves to wash seals off ice, turning terrain into a weapon and displacement into the decisive move. This is adversarial policy expressed through the environment itself. The prey’s position constrains the attacker. The attacker changes the constraint by changing the local physics. That is not a plan. It is an action-selection rule that exploits the medium, contingent on prey location, and it works or fails in the only currency that matters: did the seal end up in the water.

If you want a small, almost comically plain proof that contingency is enough, look to the spider that hunts spiders. Portia and her kin do not charge headlong. They take detours. Crucially, they choose the correct detour even after the prey has vanished from view, as though the route were selected under occlusion rather than improvised from scent and happenstance. One may argue, as sceptics do, that the rule is still an evolved routine. Fine. The point stands: a policy that selects an indirect approach under constraint, and persists with it when the quarry is no longer perceptually present, already satisfies the minimal claim. The strategic artefact is not language. It is the preservation of an alternative in the face of friction.

And now, because we are trying to stay honest, a built-in restraint. Not every group hunt is a well-orchestrated ballet. High-resolution tracking of African wild dogs suggests that group hunting benefits can arise without evidence of high-level cooperative chase strategies. That is not an embarrassment. It is a guardrail. It tells us what we should have suspected: strategic outcomes at the group level can emerge from distributed interception and additive pressure, without any individual playing a grand game. The universality claim does not require that animals behave like miniature generals. It requires something both smaller and stricter: that the system exhibits contingent policy selection under constraints, whether that policy is centrally “owned” or collectively expressed.

So in adversarial settings, across air, sea, ice, and savannah, the same footprint keeps appearing: alternative policies, constraint trade-offs, and behaviour that changes with the competitive state. That is the phenomenon we are trying to name. It is not human. It is not corporate. It is older than language.

Scope does not permit a full menagerie of corroboration, but it is worth noting how readily the same adversarial grammar reappears once you start looking for contested sensing and state-triggered switching. In the night air again, some prey do not merely flee the bat’s inference engine, they answer it, spitting ultrasound that can confound or clutter the predator’s echolocation, a kind of defensive counter-jamming that turns “who hears whom” into the decisive theatre. In daylight, other prey do not out-run pursuit so much as dissolve it: sudden pattern shifts in schooling fish, flash-expansions and contractions, the quick reweaving of the silvered geometry, can force an attacker into a costly policy change, from targeted strike to crude charge, and back again when the state permits.

Nor is interference always sonic. There are species that make the background itself their ally, switching from motion to stillness, from conspicuousness to vanishment, as though the policy were “be a thing” rather than “be a creature”. You may call it camouflage, yet functionally it is a conditional rule about what the other side may infer, and when. The point is not to heap examples. It is to show the consecution: wherever sensing becomes the bottleneck, counter-sensing blooms; wherever tempo decides, tempo is managed; wherever the opponent adapts, the policy cannot remain invariant without paying in failure.

In the human species, the same adversarial signatures appear whenever sensing and timing decide outcomes. Consider radar and radio in twentieth-century warfare: once detection became decisive, forces began to invest not only in better sensors, but in degrading the enemy’s sensors. Jamming, spoofing, decoys, emissions control, low observability. None of this requires a philosophical theory of conflict. It is simply what a competitive system does when the state of the contest shifts from “who has the bigger gun” to “who sees first, and who can prevent being seen”. The policy is contingent on the informational regime, constrained by energy, bandwidth, geometry, and countermeasure, and its consequences are measurable in detection ranges, targeting accuracy, and attrition. The functional structure is identical to nocturnal interference in the air: sensor denial as a state-triggered policy under constraint.

Or take pursuit–evasion on city streets, where the prey and predator are both human and the terrain is concrete. A police unit attempting an interception does not rely on “running faster” as its only policy. It selects among pursuit, containment, prediction, and coordination based on state: traffic density, visibility, radio coverage, escape routes, the number of units available, and the perceived risk of escalation. The pursued individual likewise switches: sprint, blend, decoy, double back, disappear into crowds, change clothing, abandon the vehicle. This is not a moral tale about cleverness. It is the same constraint logic written into a different ecology. Alternatives exist. Trade-offs bite. Policies shift with state. Outcomes follow.

If that is true in tooth and tide, the next step is uncomfortable but clarifying. The same functional signature appears in what we call politics, coalitions, and governance.

Politics, coalitions, and governance as strategic phenomena

If adversarial strategy is easiest to accept, politics is easiest to deny. We tell ourselves it is uniquely human, born of language, law, and long memory. Yet the underlying burden is older and plainer: any social system that persists must regulate conflict, allocate influence, and prevent internal friction from becoming fatal. That is governance in its most elemental sense. Not constitutions and committees, but constraint-driven control of who can do what to whom, and at what cost. In a social landscape, the scarce resource is not only food or territory. It is tolerable order to maintain the coherence of the larger strategic organism.

Begin offshore, where the water is clear enough to make you think nothing is hidden. In some dolphin populations, males form alliances that operate at more than one level: a small group will cooperate to secure access to a female, while larger alliances form to contest that access against rival alliances. What makes this strategic, under our ground rules, is not a sentimental narrative about friendship. It is the architecture of contingent support. Partners are selected. Support is deployed in specific states of competition. The presence or absence of allied bodies shifts the pay-off surface. And when the landscape changes, when rivals appear, when the balance of strength tilts, the coalition logic tilts with it. A dolphin does not need a manifesto to participate in coalition leverage. The leverage is in the structure.

Now step into the forested-fringe of primate society, where proximity is both protection and provocation. In macaque groups, conflict is not merely a private bout betwixt two combatants. There are individuals who intervene, not always on behalf of kin, but as third parties who cancel the climb of escalation and reset the terms of engagement. It is tempting to tell a cosy tale about “peacemakers”. Resist it. This is a governance function. A group that allows every quarrel to run to its bitter end pays for it in injury, stress, fractured attention, and the slow canker of coordination. Policing is a policy: intervene when conflict crosses a threshold, suppress the spiral, preserve the network. It is strategic because it is contingent, because it is costly, and because it produces observable system-level effects in stability and cohesion.

A third example sharpens the point. In chimpanzee societies, dominance is not a simple matter of body size and brute force. It is negotiated, and it is renegotiated, through coalition support. Rival males build and spend social capital, often through grooming and proximity, and they deploy that capital in the moments that matter: contests, challenges, the slow assembling of a majority. Again, avoid the theatre of intention. The mechanism is enough. Alternative policies exist. Align with the current dominant and gain immediate safety. Or back a challenger and accept near-term risk for a possible future pay-off. The choice changes with state. The consequences are legible in who wins, who loses, and how stable the hierarchy becomes.

Even in smaller societies, the political register appears the moment membership must be paid for. In a cooperative breeder cichlid, idle helpers are not merely ignored; they are sanctioned. When helpers are experimentally prevented from working, they receive increased aggression and thereupon become more submissive, as if reminded of the rent they owe for remaining within the walls. Better still, the severity of the sanction shifts with group size. That is governance, not temperament: enforcement tuned to circumstance, in service of holding the collective together.

Time would fail us to adduce every instance in which governance is not a human invention but a viability technology. In eusocial insects, for example, reproduction is not merely a biological happenstance but a policed privilege; workers suppress, harry, or eliminate unauthorised attempts, and the colony thereby keeps its internal arithmetic from becoming a riot of competing lineages. In other cooperative systems, “rank” is maintained not by constant violence but by contingent suppression, a switching rule that says, in effect, “enough”: escalate when the challenge threatens coherence, de-escalate when further conflict would cost more than it can ever repay.

And there are cases that look, at first blush, almost too bare to be called politics, yet they do the same work. Some societies impose “rent” for membership, making help and compliance the price of staying within the walls; others deploy third-party intervention that functions less as tenderness than as a circuit-breaker. In each, the scarce resource is tolerable order, and the strategic object is the social network itself, held in ward against its own fractious impulses. That the outward gestures differ is matter; that the control function repeats is the sooth.

So much for the non-human record; let us, without any special pleading, observe the same coalition logic at work in Homo sapiens. First, consider parliamentary coalition formation in polities where no party reliably holds a majority. After an election, the habitat is thick with signals: vote shares, ideological distances, reputational commitments, and the looming constraint of governing viability. Parties do not simply announce preferences. They bargain. They trade concessions. They impose discipline on defectors. They adjust their coalition policy when the balance of seats shifts, when public sentiment turns, when a scandal alters credibility, when a partner becomes too costly to carry. There is no need to flatter this with lofty nouns. It is coalition strategy as a recurrent solution to the problem of collective action under power fragmentation.

Second, consider corporate governance in large organisations, which is one of the more domesticated theatres of human politics. A budget cycle is a seasonal contest for resources. Departments form temporary alliances, not by declaring war, but by trading support across decisions. Committees act as arenas. Gatekeepers police norms. Compliance functions and HR interventions are, at their best, a form of internal conflict regulation that prevents local contests from destroying system coherence. When the environment tightens, when margins shrink or reputational risk rises, you can watch the policy switching in real time: stricter control, sharper sanctions, different alliance patterns, altered escalation thresholds. It is the same functional signature, simply expressed in a different medium.

So politics, in this sense, is not uniquely human and not inherently Machiavellian. It is a strategic response to a universal constraint: social systems must remain coherent while distributing power and suppressing destructive conflict. The outward symbols differ. The underlying structure repeats.

And once governance exists, exchange becomes possible at scale. Which creates the next strategic problem: cooperation under the perpetual risk of cheating.

Cooperation, trade, and contract stability without contract law

Once a social system can suppress its own internal fires, a new possibility opens. Individuals can do more than coexist. They can cooperate. And cooperation is not a greeting-card virtue. It is a hard problem with sharp edges. The gains can be real, but the incentives are crooked. If I can take the benefit without paying the cost, why would I not? This is the ancient riddle of moral hazard, and every stable cooperative system is, in effect, an answer hammered out against it.

That is why cooperation is not the absence of strategy. It is one of its strictest examinations.

A cooperative relationship is a corridor through a risky landscape. It may even lead to the formation of a larger, more powerful strategic organism.

Yet, it persists only if the corridor is patrolled, priced, and periodically repaired. Call it reciprocity, call it mutualism, call it exchange. The more exact description is viability engineering under repeated interaction: policies that keep value flowing while making exploitation expensive enough, often enough, to be unattractive.

One answer is partner choice. Another is punishment. A third is reputation. On a coral reef the drama plays out like a huckster’s booth. Cleaner fish run “stations” that larger client fish visit to have parasites plucked away. The exchange is real, and so is the temptation to defect. A cleaner can take a more profitable bite of client tissue rather than stick to the contractual diet. The system remains stable because the client is not a passive mark. Clients can punish and depart. Crucially, the enforcement is not blunt. It is graded. The severity of sanctioning can scale with the stakes of the relationship. After punishment, cleaners shift behaviour toward greater cooperation. This is not morality. It is contract stability as a policy: a contingent enforcement rule that preserves trade under constraint, tolerating small deviations when switching costs are high and sanctioning more sharply when the stakes demand it. An integrated set of choices, if you like, but the integration is functional, not rhetorical.

Take the same logic into a cave, into the hot, close press of a bat roost, where the constraint is not etiquette but starvation. Vampire bats live on a diet that can make a single missed meal consequential. In that regime, food sharing functions as an insurance structure. Individuals that have fed can transfer a portion to those that have not, and the pattern of giving is shaped not only by kinship but by prior exchange histories. Again, you do not need to inflate this into human-style intention. The mechanism is sufficient. Under a harsh constraint, a contingent transfer policy stabilises a repeated relationship that reduces the variance of individual failure.

Then there is the rat work, persuasive precisely because it is spare. In rat cooperation experiments, reciprocity is not merely a reflexive “you scratched my back, I scratch yours”. It can be sensitive to value. Not merely “you helped me once, therefore I help you once,” but “the quality of what you did for me changes what I do for you.” That is a harder test than warm stories about social bonds. It suggests a policy that integrates partner history, distinguishes value, and adjusts repayment accordingly. This matters because it rules out the most convenient dismissal, the one that says, “It’s just mimicry.” Value-sensitive reciprocation is a mechanism for maintaining cooperation when contributions differ, when free-riding is possible, and when the system needs a way to calibrate repayment to what was actually provided. That calibration is a policy. It is contingent. It has consequences. And it functions as a contract analogue without contract law.

A third mechanism is the quieter one, the one humans often overlook because it does not look like punishment. It is the simple, relentless discipline of switching. In repeated interactions, the option to withdraw, to choose a different partner, to withhold future benefits, is itself an enforcement tool. Cooperation often stabilises not because individuals become saints, but because the landscape makes defection expensive in the long run. When the state changes, when a partner’s reliability is inferred to decline, cooperative policy shifts. Help less. Help differently. Help elsewhere. This is not sentiment. It is adaptive exchange under constraint.

Nor is this confined to creatures with eyes that meet yours. In the legume–rhizobium bargain, the plant withholds payment from underperforming nodules. In the mycorrhizal trade, the plant allocates more carbon to fungal partners that deliver more nutrients. It is almost drearily commercial: performance measured, rewards varied, cheats starved of upside. Call it physiology if you like. Functionally, it is the same stabilising logic you would expect in any repeated exchange where quality can be observed and compensation can be made conditional. Even in microbial commons, where sentiment is a nonsense-word, cooperation can persist because production, access, and exclusion are structured such that cheats are, oft times, made to thole the consequences of their thrift.

We could wander, whilom and now, through a whole market of non-human contracts, and still not exhaust the catalogue. There are inter-species bargains where enforcement is not a court but a constraint: partners who deliver more are fed more; partners who shirk are quietly starved of future gain. There are mutualisms where punishment is less a dramatic reprisal than a calibrated withdrawal, a policy of switching that makes defection costly over the only horizon that matters, the repeated game.

Now a guardrail, because universal claims rot when they become sentimental. Not every cooperative-looking pattern is contingent exchange. Sometimes group-level benefit arises from by-product mutualism, or from distributed effects that do not require high-level coordination. In other words, co-operation does not automatically form a larger strategic organism. There is no contradiction. Cooperation can happen between discrete strategic organisms within an ecosystem without the ecosystem being a strategic organism. (This is very often the case in that collective the human primates like to call “a company” or “a firm”). Chimpanzee hunts, for example, can display catalytic dynamics where a few initiators shift the odds for everyone, and careful authors explicitly warn against over-reading cognition into the pattern. That is not a nuisance. It is a boundary condition. Where partner history is not used, where switching is absent, where sanctions do not vary with state, you may still have cooperation, but you have not yet earned “strategy” under your own ground rule. Natheless, where repeated interaction, cheating risk, and enforceable contingencies coincide, the exchange begins to look uncannily like contract stability, even without contract law.

But enough of reef and roost; the same constraint-logic appears, in plainer clothing, in the exchanges of Homo sapiens. First, consider trade credit between firms. A supplier extends terms, not because it is charitable, but because it is selecting a policy under uncertainty in a repeated relationship. Credit limits change with inferred counterparty risk. Payment terms tighten when liquidity tightens. Discounts and penalties are tuned to discourage delay and default. Whole enforcement ecologies arise around this: scoring, insurance, factoring, escalation, collections, legal escalation. What looks like “finance” is often simply cooperation engineered to remain stable under cheating risk and asymmetric information.

Second, consider open-source software communities, a modern laboratory for cooperation without a central sovereign. Contribution is voluntary. Defection is cheap. And yet high-quality cooperation persists, but only because the system builds contract substitutes: reputation via commit history, maintainers as gatekeepers, norms about review and testing, the power to accept or reject changes, and the ability to fork when governance fails. When trust is high, the policy is permissive. When the threat of sabotage rises, the policy hardens: stricter reviews, tighter permissions, slower merges. It is the same functional signature in a different ecology: state-dependent policy selection under constraint, designed to keep cooperation from collapsing.

So cooperation is not a “soft” domain that sits outside strategy. It is where strategy becomes most exacting, because it must extract mutual gain while continuously policing the boundary between contribution and exploitation. But cooperation is never only about what you do. It is also about what others believe, or infer, you will do next. Which brings us to the most neglected strategic domain in business, even though we practise it daily: information strategy.

Information strategy (signals, deception, reputation, interference)

If cooperation is the corridor, information is the lantern. It influences who steps forward, who hesitates, who flees, who fights, who trusts, who cheats, who bargains. And because information changes behaviour, information becomes contested. It is not merely shared. It is shaped. In many systems, the decisive strategic move is not a strike, or a trade, or a migration. It is an alteration of what another agent can infer.

Begin with the simplest unit: a signal. A signal is not a speech. It is any cue that modifies another’s policy. The strategic question is whether the cue is reliable, and if it is not, whether the receiver can discriminate it. This is why deception emerges so readily. In the Kalahari, fork-tailed drongos exploit the listening habits of other foragers. They produce alarm calls that trigger a predictable response: food is dropped, attention is seized, movement is redirected. The drongo does not need to “lie” in a moral sense. The functional mechanism is enough. A signal is deployed that shifts the receiver into an expensive defensive policy. The sender then harvests the benefit. And because the receiver population is not inert, countermeasures appear. Targets learn. They habituate. They discount. Drongos respond by varying the calls they use, keeping the deception effective across time. That is the informational arms race writ small, in dust and thorn and stolen morsels.

Now shift from sound to scent, from a momentary cue to a lingering signature. In spotted hyenas, social information is carried in odour. What makes this particularly instructive is that the signal is mechanised. Odour profiles are shaped in part by microbial communities. A stable chemical signature becomes a kind of identity token, a way for others to infer who is present, where they belong, and how they should be treated. This is not sentiment. It is an information channel with real consequences for access, conflict, and affiliation. The point is not that hyenas “care” about reputation in the human way. The point is that the system uses a persistent signal that regulates social policy, and that signal has a causal substrate, not a metaphorical one.

Deception is not confined to opportunistic theft. It can be structural. Consider brood parasitism in birds. A cuckoo’s problem is brutally specific: offload parental investment onto another species. The host’s problem is equally specific: do not raise someone else’s offspring by mistake. In that duel, the battleground is inference. Egg mimicry is not a decorative flourish. It is a deliberate move in signal space, shaped by selection pressures that reward successful misclassification. Hosts respond with discrimination policies: egg rejection, nest desertion, increased vigilance. Parasites respond with better mimicry, altered timing, or other tactics. This is strategic behaviour across generations, but the signature remains the same: contested information leads to policy selection, and policy selection reshapes the competitive landscape.

Even when no deception is involved, signalling can still be strategic because it can be costly, and cost is what makes credibility possible. Stotting in gazelles, for instance, has long been discussed as a signal that influences predator decisions: it makes the prey appear less worth pursuing. Whether you interpret it as honest advertisement of fitness or as a deterrent display, the functional core is the same. The signal is not merely expressive. It is instrumental. It shifts the predator’s policy by altering inferred prey state. It is a move in the belief-state of the other side.

And then, because the world is rarely courteous, there is interference again, but now in the informational domain rather than the kinetic one. Certain animals eavesdrop on alarm calls and exploit them. In social species, audience effects shape how signals are deployed: the same aggressive display, the same call, the same posture, can change depending on who is watching, because the real recipient is not always the immediate opponent. Information strategy is often triadic. The “message” is aimed at the network.

It would be easy, and a little dishonest, to make this section a parade of cunning. Better to note how broad the evidential palimpsest already is, even when we refuse to read mind into it. There are cephalopods whose skins are not merely coloured but instrumental, a live surface that can switch from proclamation to concealment, from “I am here” to “I am weed” as the state demands. There are insects whose entire life-histories are built on the manipulation of receiver inference, not as a flourish but as survival arithmetic: mimicry that buys a margin, camouflage that buys a moment, timing that buys access.

And then, since you asked for the occasional breaking of the mould, permit one brief divagation. Consider how quickly “reputation” becomes physical once the signal persists. A scent-mark, a chemical profile, a patterned display, an acoustic signature, all become identity tokens that other agents use to select policies of approach, avoidance, affiliation, or attack. The signal is no longer an episode; it is infrastructure. The strategic burden shifts from “can I deceive once?” to “can I remain credible across time, under scrutiny, while rivals learn and discount?” That is not a moral question. It is an ecological one. The moment inference governs action, the control of inference becomes a primary theatre.

With that in view, we may turn to Homo sapiens, not as a special case, but as one more animal that builds elaborate machinery for shaping belief. First, consider commercial advertising and branding, stripped of romance. A brand is a persistent signal that aims to alter customer inference under uncertainty. It compresses a promise into a cue. The strategic problem is credibility. The signal must be reinforced by costly actions that are hard to fake: warranties, consistent quality, long-term investment, visible sacrifice of short-term margin. When the environment tightens, when trust erodes, when counterfeits and low-cost entrants proliferate, the species responds with signal hardening: certification marks, traceability, legal enforcement, reputation systems. This is not culture. It is information policy under constraint, designed to steer behaviour in a competitive market ecology.

Second, consider organised deception in conflict, which is simply information strategy with higher stakes. Armies use decoys, feints, camouflage, misinformation, and electronic warfare to shift an opponent’s inference loop. The aim is not only to conceal capability, but to induce the enemy to act wrongly: allocate force to the wrong place, commit too early, hesitate too long, strike the empty field. Receivers adapt with verification protocols, redundancy, and hardened sensing. Senders respond with more layered deception. Again, the logic is plain. If action follows inference, then shaping inference becomes a primary theatre of strategy. It is the same structure we saw in drongo alarms and cuckoo mimicry, simply executed with more elaborate apparatus.

Information strategy, then, is not a decorative add-on to “real” strategy. It is often the core. It is the control of what others can know, what they can trust, and therefore what policies they select under constraint.

And because inference is always bounded, because the future is always uncertain, the next strategic burden follows naturally: how systems manage time through buffering, redundancy, and pre-positioning.

Logistics, buffering, robustness (strategy as time-binding)

After conflict, after coalitions, after exchange, after signalling, there remains a quieter truth. The future is not a date. It is a constraint waiting its hour. In uncertain worlds, survival is less about the brilliance of any single move and more about whether you can keep acting when the weather turns, when the prey vanishes, when the rival arrives, when the stream runs dry - or floods. This is the logistics dimension of strategy: the discipline of time-binding. It is where a system converts surplus into resilience, and resilience into option.

Start with caching, because caching is strategy you can hold in your hand. Western scrub-jays store food in places where it will be available later, and the pattern of storage can shift with anticipated future need. That is not a mere hoarding reflex. It is allocation contingent on expected state. The bird distributes resources across time and location, and in doing so it changes the future action space. What looks like a small, ordinary act of hiding a nut is, functionally, a policy about scarcity risk.

Now add a second layer: security. A cache is only a buffer if it remains yours. Scrub-jays adjust their behaviour when watched. They will relocate stores, recache, change locations. The world state includes observers. The policy changes when pilferage risk rises. It is difficult to ask for a clearer, cleaner signature of strategy. Alternative policies exist. The trade-off is real. Time and effort spent moving caches is time not spent feeding. But the cost is paid when the risk of theft makes the original policy too brittle to bear.

Move from birds to mammals and the logic becomes heavier, more bodily, more expensive. Many species store energy in tissue. Fat is inventory. It is not glamorous, but it is a buffer against volatility. Under conditions of seasonal scarcity, migration, or reproductive demand, organisms shift their allocation policies between immediate consumption and storage. Even without telling any sentimental story, you can see the strategic structure: the organism has an internal state, and it adopts different allocation policies under different constraint regimes. A stable store is not a plan. It is an engineered capacity to continue acting when the environment stops being generous.

Even insects, in their own austere way, rehearse the same logic. Ant colonies store food, manage brood allocation, and regulate labour distribution. Their buffering is not a warehouse with spreadsheets. It is distributed across bodies, chambers, and behavioural roles. Under threat, under shortage, under pathogen exposure, the colony shifts allocation rules: who forages, who stays, who interacts with whom. Logistics is not only about stores. It is about controllable flow. A colony’s ability to reroute effort and isolate risk is a kind of operational resilience, expressed as policy switching under constraint.

Now take the same logic into the long drought-gait of elephants. In dry seasons, when water becomes a governing constraint rather than a convenience, elephants adjust movement and grouping policies in ways that are plainly state-dependent: routes change, timing shifts, risk tolerance tightens around calves, and herds can favour known water sources over speculative search. The behaviour is not merely locomotion. It is logistics under scarcity. A waterhole is not a “nice place”. It is a node in a survival network, and the path taken to it is a trade-off among heat, distance, predation risk (yes Elephants are well known to avoid poachers), social cohesion, and the simple arithmetic of depletion. If the state changes, the policy changes. That is the strategic footprint we have been tracking, expressed at the scale of kilometres and days rather than minutes.

Now add the sharpest kind of constraint: disease. Some primates, including mandrills, show avoidance behaviours linked to infection cues. The contact network is not fixed. It is rewired in response to inferred threat. This is robustness as topology. Under certain states, the best logistics policy is not “more movement” but less. Not “more interaction” but modularity. It is the same principle a ship’s designer uses when building watertight compartments. Contain the failure before it becomes systemic.

So the animal record gives us multiple forms of time-binding: caching, recaching, body storage, labour reallocation, and network rewiring. The outward forms differ because the substrates differ. The functional invariants repeat: allocate surplus into buffers, protect buffers, and change allocation rules when the risk landscape changes.

If one wished to press the point further, one could do so without leaving the province of simple stores. There are small mammals that lay up winter in a thousand hidden larders, and must then solve the second-order problem you already noted in birds: not merely caching, but guarding, relocating, and deceiving, because a buffer is only a buffer if it persists. There are creatures that build haypiles, chambers, granaries, combs; each is a wager that present surplus should be turned into future option, and each invites the same strategic question: how much slack can you afford before slack itself becomes the ruin, by weight, exposure, or theft?

And robustness is not only inventory. In some systems it is architecture, and in others it is rhythm. A colony may not store “things” so much as store role-flexibility, keeping labour in equipoise such that it can be rerouted when drought, predation, or pathogen makes the old flow-pattern untenable. The mechanism is unromantic and therefore persuasive: time-binding is policy, and policy switches when the risk landscape changes.

Let us now watch our own kind perform the same time-binding tricks, but with warehouses, budgets, and protocols in place of burrows and comb. First, consider inventory and redundancy in supply chains. When volatility rises, the species shifts its policy. It holds more safety stock. It dual-sources. It builds slack into schedules. It moves from “just-in-time” toward “just-in-case” in specific nodes, not everywhere, because slack is costly. The switching is visible and measurable. It is driven by state inference about lead times, disruption probabilities, and dependency fragility. The behaviour is not a moral choice between prudence and waste. It is logistics as strategy: buying the ability to keep acting when the environment becomes adversarial.

Second, consider civil defence and emergency provisioning. In threat regimes, human polities pre-position supplies, maintain reserves, harden infrastructure, and design protocols for continuity. When the perceived threat state changes, budgets shift, drills proliferate, reserves are replenished or allowed to decay. Again the pattern is not mysterious. A system that expects volatility invests in buffers. A system that believes the world is stable strips them out to gain efficiency. Both are policies. Both have trade-offs. Both have consequences that become painfully visible when conditions change.

Logistics, then, is not the dull cousin of “real strategy”. It is strategy’s time axis. It is how a system refuses to be held hostage by tomorrow. In the parlance of the SO(X) theory it is a vital part of closing the resource cycle.

And once you see buffering as strategic, movement itself changes character. Navigation becomes more than travel. It becomes search. It becomes allocation of attention across a landscape of uncertain rewards. That is where we go next.

Mobility, navigation, exploration–exploitation

Once you accept buffering as strategic, movement stops being a neutral act. Travel becomes search. Search becomes a wager. Every step is an allocation of effort across a landscape whose rewards are patchy, perishable, and often concealed. The strategic burden here is not “going somewhere”. It is deciding, under constraint, how long to exploit what you have found, when to abandon it, how far to range, and how quickly to commit when time turns hostile. Exploration and exploitation are not academic abstractions. They are the pulse of any foraging system, whether the forage is nectar, prey, shelter, mates, or information.

Some of the cleanest evidence comes from insects precisely because the mechanisms are legible. In house-hunting ants, collective movement is not a dreamy drift. It is governed by thresholds. A colony scouts, evaluates, recruits, and then commits when enough confirmations accumulate. Critically, those thresholds are not fixed. Under harsher conditions, colonies lower the bar. They commit faster, accepting a higher risk of error in exchange for speed. Under safer conditions, they demand more evidence, trading speed for accuracy. That is the signature we set as our boundary: alternative policies, tuned to state, with consequences. It is a speed–accuracy trade-off expressed as a switching rule, written into the colony’s decision architecture.

Now move from commitment to route, from choosing a home to choosing a path. Bumblebees, and other foragers, develop repeatable “traplines”, stable routes that link multiple resource sites. The important point is not that a bee can remember. The important point is that the route is a policy. It can be improved, and it is improved, through experience, as the forager refines sequence and distance to reduce cost. When the environment changes, when a flower patch fails, when competition rises, when wind makes one corridor expensive, the route adapts. What you observe is neither random wandering nor rigid habit. It is conditional optimisation in a shifting field.

Hummingbirds offer a similar lesson, sharper in its geometry. Nectar rewards replenish on a schedule. Visit too soon and you waste flight for little gain. Visit too late and a rival has harvested your profit. The effective policy, therefore, is not merely “visit flowers”. It is “visit on this cadence, in this sequence, with this revisitation interval, given the refill rates and the local competitive pressure”. The bird’s movement becomes a metronome tuned to ecology. Again, this is not a story about miniature accountants. It is a functional response to a constraint: energy is costly, and the reward landscape is time-dependent.

Then there are navigators whose problem is not patchwork foraging but long-distance return. Homing pigeons, migratory birds, sea turtles, salmon. The interesting strategic question is not whether they can orient. It is whether they can switch navigation policies when particular cues become unreliable. In fog, in geomagnetic anomalies, in unfamiliar terrain, in high winds, the effective policy cannot be a single invariant routine. Where cue quality degrades, reliance shifts. Landmark use gives way to other modalities. The system weights whatever remains dependable. That reweighting is a form of inference, and it is expressed behaviourally as route correction, timing changes, altitude adjustments, and altered stopover choices.

At the mammalian end of the spectrum, the same logic appears with more visible risk. A troop moving through a landscape is not only chasing food. It is managing predation threat, social cohesion, and energetic cost. Movement policies change with infant presence, with the proximity of rivals, with the availability of cover, with water stress. The route is not just a line on the ground. It is a compromise between competing constraints, revised as state changes.

We might also, had we the room, follow the same logic into the long-distance and the mathematically strange. There are foragers whose search paths resemble a kind of disciplined wander, neither random nor rigid, as though the policy were tuned to a landscape where reward is sparse and uncertainty is honest. There are desert navigators that return home by integrating their own movement, keeping a running reckoning when landmarks are absent, and adjusting when wind, slope, or detour makes the body’s own odometer unreliable. There are migrants that do not merely “know where to go” but appear to reweight cues when one channel becomes treacherous, shifting reliance as fog, magnetic noise, or unfamiliar coastlines degrade the signal.

The important point is not virtuosity. It is contingency. The route is a policy, and the policy changes when the state changes: when time pressure rises, when predation risk thickens, when information quality improves or collapses. In that sense mobility is not travel. It is inference made legible on the ground.

And so, with no change in the underlying burden, we come at last to Homo sapiens, whose maps and machines merely make the same switching rules louder. First, consider urban movement. Humans do not merely move from point A to point B. They switch routing policies based on inferred congestion, surveillance risk, time pressure, and energy cost. When the environment is calm, they optimise for convenience. When threat rises, they optimise for concealment or speed. When information quality improves, via maps, signals, reports, they change route selection accordingly. In a modern city, the decisive strategic resource is often not horsepower but inference: who can predict congestion, who can anticipate chokepoints, who can choose an alternate corridor before the herd arrives.

Second, consider the human species in its corporate colonies, engaged in the long foraging of innovation. Research and development is exploration. Scaling a proven product is exploitation. Firms that survive do not pick one forever. They switch. When margins thin, when a market saturates, when a technology shifts, they push resources toward exploration. When uncertainty is costly, when capital tightens, when a path proves reliable, they swing back toward exploitation. The signature is again the same: alternative policies, constraint trade-offs, and state-dependent selection with observable consequences in portfolio outcomes.

Mobility strategy, then, is not about being restless. It is about searching without starving, committing without blundering, and learning without losing coherence. It is the art of moving through uncertainty without being consumed by it.

And once a system begins to navigate in this way, it eventually confronts a deeper possibility: instead of merely moving through the landscape, it can reshape the landscape, or extend its own capabilities, so that the next search is cheaper. That brings us to technology and niche construction.

Technology and niche construction, or how capability changes the game

Up to this point, strategy has looked like the selection of moves within a given game: pursue, evade, bargain, buffer, route. But living systems do not always abide by the game they inherit. Sometimes the most consequential strategic act is not choosing better moves. It is enlarging the set of moves that are possible at all. Capability changes the game. Constraint is not merely endured; it is edited, re-priced, and, in the boldest cases, re-made.

The cleanest way to say this, without romance, is that a tool is an external object or technology used to alter the cost, reach, or reliability of an action. Under that definition, the poetry belongs to affordance, not to “intelligence”. A New Caledonian crow fashions a hooked tool from a sliver of plant, then coaxes larvae from the stubborn fastness of wood. You can almost feel the grain resisting, hear the small scrape and fret of fibre. The hook is a slight thing, a hand-span of wit made hard. Yet it changes the feasible set. What was sealed becomes yielded. What was costly becomes cheaper. The crow is not merely choosing where to forage. It is choosing to change the foraging problem itself.

Now shift to a creature that does not look, to human prejudice, like a tool user at all, and therefore does honest work for our argument. The coconut-carrying octopus collects coconut shells, carries them across open ground, and later assembles them as shelter. The sequence is the telling detail, the very marrow of it. Acquisition. Transport. Delayed deployment. Present vulnerability accepted for future protection. The shell is not a hiding place stumbled upon in a pinch. It is a portable fortification, borne through risk and unfolded when state demands it. If you want the functional signature in one line, it is this: capability is being moved through time.

Primates offer further instances, and they do so with a pleasing plainness. Chimpanzees use sticks to fish termites, stones to crack nuts, leaves as sponges. Each behaviour is a small engineering intervention, a deft rede of matter into method, altering the cost structure of acquiring resources. The stick makes the insect accessible. The stone makes the shell yield. The leaf makes water gather where none seemed to be. The object is not a symbol. It is a lever on the world. Again, we do not need to smuggle a boardroom inside a skull. We need only observe that the policy now includes an external implement, and that inclusion opens value channels that bare limbs could not.

Then there is niche construction, which is tool use made edifice, and architecture made strategic. A beaver does not merely seek a safe place. It makes one. A dam turns running water into held water, a wild rush into a steadier store. It moderates drought and dilutes danger. It reshapes local ecology. It changes what is predictable. The point is not that the beaver is a visionary hydrologist. The point is that the system pays a present cost to rewrite future constraints, so that the next decision is taken on gentler ground.

Termites and ants do similar work with different substances and scales, like masons of mud and breath. A mound is not an ornament. It is a microclimate, a ventilation logic in earth and saliva, a stable internal atmosphere carved out of a hostile external one. A nest is a device for constraint management. It enlarges the range of conditions under which the system can remain coherent. Even biofilms, in the microbial world, build a matrix that changes diffusion, exposure, and vulnerability. Cells that would be swept away become cells that persist, not because the world became kind, but because the local world was remade. Do we claim that the ant is strategic? Enter again the construct of the strategic organism - SO(X) - a theory which admits the ant colony to qualify as a strategic organism.

Now, two examples from Homo sapiens, treated as observations of an unusually artefact-laden primate. First, consider the invention and adoption of long-range sensing and strike technologies in conflict. A drone, a satellite, a secure network. Each is not merely an instrument. Each expands the action space by decoupling sensing and acting from bodily proximity. It changes tempo. It changes concealment. It changes the economics of risk. Countermeasures bloom in response, because the environment has been altered. Strategy here is not the strike itself. It is the adoption of a capability that re-sculpts what is feasible and what is foolish for all parties.

Second, consider platform-building in commercial habitats. A platform is niche construction in market space. It builds an environment in which other agents act, and it does so by providing standards, interfaces, enforcement, and distribution. It makes certain behaviours cheap, others costly. It attracts complements and disciplines defection. This is not a buzzword about “business models”. It is ecological engineering, except the ecology is contractual and computational rather than botanical.

So capability extension and niche construction reveal a deeper layer of strategy. Strategy is not only the selection of policies within constraints. It is also the creation of new constraints, new affordances, and new futures. Some systems do not merely play games. They change them.

At this point, the argument is no longer about animals, and it is not even about humans. It is about the underlying phenomenon. Which brings us to the hinge: why human business talk makes strategy look more complex than it is, by confusing the artefacts we build with the function they are meant to serve.

The hinge, or how the human species mistook its tools for the thing

By now the pattern should feel almost stubborn, even stony. Different bodies, different resolutions of strategic organism - yes. Yet the same footprint keeps appearing: alternative policies, binding trade-offs, state-dependent selection, consequences you can see without rummaging in anyone’s head. That is the phenomenon.

So why does the human species speak of “strategy” as if it were a mist that cannot be held, a word that means everything and therefore nothing? Why does the conversation so often turn shaggy, scholastic, and recondite, full of quarrels about definition, followed by the consoling verdict that strategy must simply be “too complex” to define? I think the arrow points the other way. The phenomenon is not what became baroque, nor the theory that describes it. Our discourse became convoluted. And it became that for a reason that is neither mystical nor malicious. It is mechanical, almost inevitable, given how we live.

Humans are a coordination-heavy animal. We do not scale by instinct alone. We scale by artefact. We require objects that allow many minds to align on a shared policy: plans, budgets, scorecards, operating models, rituals, narratives, decks. These are not worthless gewgaws or gimcracks. They are technologies of alignment, instruments of assent, scaffolds for shared action. They help a large group behave as if it had one policy rather than a thousand competing impulses - this is necessary for the human herd to form a larger strategic organism. But an artefact is not the phenomenon it serves. A nest is not the bird. A pheromone trail is not the ant. A map is not the land.

The trouble begins when we promote the tool - the product - into the definition, when we forget the difference between means and meaning. We start to say strategy is the plan, or the framework, or the transformation programme, or the cascade of objectives, or the culture initiative. Then we become surprised that no single definition holds. Of course it does not. Each artefact is designed to solve a particular local burden: alignment, execution discipline, competitive positioning, resource allocation, belief-shaping. Each artefact catches a slice. None can catch the whole. Yet each is marketed, taught, and repeated as if it were the thing itself. The result is a kind of definitional driftwood, the residuum year after year, until the shoreline looks like wreckage and we call it complexity.

You can see this drift in the way the human species rewrites its slogans, polishing them like coins until the shine becomes an encrustation that hides the underlying metal, and then spending them as if they were the whole treasury. One season, it announces that strategy is “where to play and how to win”. Another season, it offers a tightened version: “an integrated set of choices that compels desired customer action.” Both lines are neat. Both are useful in their proper place. Both gesture at genuine features of competitive life. Yet notice the quiet tell. The phenomenon has not changed. Only the packaging has changed. A lion does not issue a revised definition of hunting in late Q4. Strategy, as phenomenon, does not wobble because our sentences do. Our opinions wobble. Our products wobble. Our artefacts wobble. A thousand perspectives may describe a single phenomenon, yet the perspective is just that. It is not the phenomenon. That churn is better explained as product-shaped lenses than as evidence that the underlying thing is inherently undefinable.

A full rehearsal of non-human “engineering” would quickly become a small book of its own, so I will only gesture at the further reaches. Some mammals use stones as anvils and hammers, not as toys but as cost-cutters that turn hard shells into accessible calories. Some marine hunters wield tools of their own, whether as carried objects, chosen substrates, or improvised barriers that change the prey’s feasible escapes. Even fish, in their different element, can sometimes appear to edit the game by exploiting the physics of water as a medium, turning distance, refraction, or pressure into a lever on what is reachable.

And niche construction, once you see it, is everywhere, like a quiet masonry. It ranges from simple shelters to elaborate climate machines, structures that make a local world more predictable than the outside world has any right to be. One is tempted, briefly, to be carried away by the resemblance to our own habitations: the mound that breathes, the matrix that binds. But the lesson is not that termites are little architects in waistcoats. It is that the strategic move can be the creation of new constraints, new affordances, new stability conditions, so that tomorrow’s choices are taken on gentler terms than today’s.

Now, as has become our pattern, two human examples described with the same cold zoology we used for dolphins and drongos. First, consider the consulting marketplace. “Strategy” exists as a menu category, with branded methods, templates, certifications, and deliverables that can be sold, repeated, and audited. In that ecology, novelty and differentiation are rewarded. A reusable artefact is easier to transact than a careful account of an underlying function. So artefacts multiply, each with its own vocabulary, each with its own claim to completeness, each shading the meaning of “strategy” toward whatever it sells. The word stretches. The arguments thicken. Confusion is not a bug. It is a predictable by-product of how the market selects.

Second, consider business education and executive ritual as a transmission system. What is taught must fit into modules. What is examined must fit into criteria. What is remembered must fit into a line. So strategy becomes a procession of frameworks and phrases, clean enough to travel, pliable enough to apply, flattering enough to repeat. Graduates become fluent in the dialect while remaining oddly uncertain about the underlying object. They can recite artefacts. They can defend artefacts. They can argue about artefacts. And because the artefacts disagree, they infer that the phenomenon must be irreducibly complex.

This is the hinge. The complexity objection is not necessarily a deep truth about the world. It may be a symptom of category error. We have mistaken the tools of coordination for the function those tools were meant to express. We have called the artefact the phenomenon, then wondered why the phenomenon appears fragmented.

Reverse the mistake and the concept becomes smaller, not larger. A definition falls out that is not “corporate”, not “human”, not product-shaped. A definition you can recognise in bats and beavers and bees, and also in boardrooms, precisely because it is about what systems do under constraint.

That is the definition I formalise in QCEA and SO(X).

Synthesis, or the small hard kernel beneath the frippery

If we stop mistaking our instruments for the thing itself, the definition becomes almost disconcertingly small. Not because it is vague, but because it is tight. It names the constraint set that makes strategy real, and it draws a boundary that can fail.

Strategy is system-level coherence-preserving inference-and-action under three binding pressures: entropic drift, irreducibly incomplete information, and scarcity of usable resources. In operational terms, it is a system selecting among genuine alternative policies based on what it can infer about state, while the world is actively wearing it down, hiding parts of itself, and rationing the very inputs required to act.

That is the universal object. It is not a plan. It is not a position. It is not a programme. It is not a slogan with good bones. Those are artefacts the human species uses to coordinate many minds around a policy. Useful often, yes. Definitional, no.

Notice how the constraints keep the definition honest. If information is complete and reliable, you can optimise by rule and forget the rest. If resources are abundant, you can blunder and survive, and call it “learning”. If entropy does not bite, if the world does not drift toward disorder and depletion, you can repeat yesterday’s routine forever. Strategy only becomes necessary, and only becomes visible as a phenomenon, when all three pressures press at once: the world erodes your coherence, your map is always partial, and your capacity to act is rationed.

This is why the universality objection is, in a curious way, backwards. The more you believe those pressures recur across domains, the more you should expect the same functional invariants to recur - and they do. Strategy becomes universal not because every system looks the same, but because the burdens that force strategy are the same: uncertainty that will not yield, constraints that will not relax, and drift that will not stop. Different bodies. Same weather.

Now, why do QCEA and SO(X) look “complicated” at first glance? Often because we forget the rule of resolution and we become easily overwhelmed in the zoological and biological examples in this piece because we forget to adjust our resolution. Are we talking about the ant colony? Are we talking about the chimpanzee as a strategist within the troop, and looking at the strategy of that individual? Or are we looking at the strategy of the troop which may behave as a strategic organism? Or are we looking at a species as an intergenerational strategic organism - which is permitted and, in fact, enabled by the theory.

You see - when we fail to adjust resolution, the examination of the inter-species empirical record very quickly feels like “I’m lost, everything feels like strategy”. Zoom in and out to the correct resolutions in each case and the functional invariants not only become clear but they become simple. We see that strategic organisms exists at different resolutions, yes but they all follow the same 18 fundamental laws of strategy. To describe these laws precisely though requires more than the comfortable slovenliness of imprecise words. QCEA is simply that constraint grammar made explicit. It says: if you want to talk about strategy as something real, you must talk about uncertainty, complexity, entropy, and adaptivity as first-class citizens, not as rhetorical garnish. QCEA is not an add-on. It is the minimal machinery you need to describe why a system must infer, act, update, and switch, rather than merely execute.

SO(X) names the matching kind of system: the strategic organism. Not “organism” in the sentimental sense of something warm and biological, but in the functional sense of a coherent agentic system, of whatever substrate, that can keep making choices under those joint pressures. SO(X) is the thing that does not collapse into brittle routine, that does not suffer ossifications, when its environment shifts, its information is compromised, and its inputs tighten. It remains viable by continuing to cycle inference into action, and action back into inference, without losing itself.

Seen in that light, QCEA and SO(X) are not complexity theatre. They are compression. They take the recurring patterns we have walked through, the jamming in night air, coalition leverage in social networks, exchange stabilised by sanction and switching, signalling wars, buffering against volatility, thresholded movement decisions, capability extension and niche construction, and they refuse to treat them as separate curiosities. They treat them as different manifestations of the same underlying grammar. QCEA names the grammar. SO(X) names the kind of system that can speak it.

Two brief field-notes from Homo sapiens, treated as one more species among many. First, watch what this species does when the environment turns sharp and the future becomes less a calendar than a cudgel. It stops trusting single policies. It seeks option. It diversifies supply, buffers inventory, hardens infrastructure, rewires networks, tightens governance, shifts from speed to accuracy or from accuracy to speed depending on threat. In short, it behaves like a system under entropic drift, incomplete information, and scarce resources. The artefacts differ. The functional signature is the same.

Second, watch what the same species does when it tries to explain what it is doing. It produces artefacts: decks, models, slogans, plans. Often necessary, sometimes splendid. Then it quarrels about those artefacts as if they were the phenomenon, and it wonders why strategy seems to change every season. QCEA and SO(X) are an attempt to step beneath that quarrel, back to the thing that does not change with fashion: coherence-preserving inference-and-action under constraint.

So the response to both objections is, in essence, the same reversal. Universality is not indulgence here. It is the economising move. And QCEA/SO(X) is not complication for its own sake. It is the minimal structure on the far side of complexity: the fewest moving parts that still explain why the same strategic footprints keep appearing across wildly different worlds.

Now we should do what serious people do when they believe they have simplified something: try to break it. There are fair objections, and they matter. Let’s meet them cleanly, without theatrics, before we close.

Pre-emptive objections, kept tight and honest

At this point a sensible reader has at least three objections, and if we do not meet them directly the piece becomes mere performance.

The first is anthropomorphism. “You are telling animal stories and smuggling in human intention.” No. I am being deliberately stingy about minds. I am not claiming language, conscious long-horizon planning, or moral categories. I am using a functional test: do we observe state-contingent policy selection under the QCEA constraint set, with consequences? Where the evidence is equivocal, we should say so. Where behaviour looks like a fixed routine, we should not call it strategy. The whole point of the ground rules was to keep the theory from becoming a flattering projection.

The second is the “everything is strategy” trap. “If all adaptive behaviour counts, the word becomes useless.” Agreed. That is why the definition has a boundary with teeth. Strategy requires contingency, real alternatives, binding trade-offs, and observable consequences. Reflexes are not strategy. Fixed action patterns are not strategy. Routines that fire the same way regardless of state are not strategy, even if they are effective. Strategy is the special case where a system can do otherwise, and does otherwise, because the state and constraints demand it. But it warrants repetition. Strategy is something that the system does - hence the construct of the strategic organism. When feeling confused, ask if you have dialed in at the correct resolution - are you looking at the strategic organism or its constituent actors?

The third is the “it’s just heuristics” dismissal. “You’re not discovering a universal phenomenon, you’re describing simple rules.” Fine. But notice what that concedes. A heuristic that selects actions based on inferred state under scarcity, incomplete information, and drift is still a policy. The question is not whether the rule is simple or complicated. The question is whether the rule is contingent, constraint-aware, and outcome-bearing. In fact, if a small heuristic explains a wide empirical footprint, that strengthens the parsimony wager rather than weakening it. Simple on the far side of complexity is not an insult. It is the whole ambition.

One more objection is worth naming because it often hides inside the word “context”. “Strategy is context-specific, therefore a universal definition is impossible.” Context matters. Of course it does. Context is the state. The universal claim is not that everyone should do the same thing. The universal claim is that across contexts, viable systems face the same class of constraints, and therefore their strategic behaviour has the same functional structure. Universality here is about the grammar, not the sentences.

Which brings us to the close.

If you have read this far, you deserve the final move to be simple.

Strategy is not a corporate ceremony. It is not a managerial mood. It is not a branded framework with a tidy quadrant. It is a phenomenon that appears wherever a system must keep itself coherent while the world degrades its options, hides information, and rations the resources required to act. That is why the universal definition is not hubris. It is housekeeping. It is the smallest hypothesis that fits the widest footprint.

And it is also why the second objection, that QCEA and SO(X) are “too complex to be useful”, misses the direction of travel. They are not extra adornment. They are an attempt to say, with some discipline, what keeps recurring across domains: coherence preserved by inference-and-action under entropic drift, irreducibly incomplete information, and scarcity. In other words, the minimal structure on the far side of complexity.

So here is the practical takeaway, not as a sermon but as a field rule.

The next time someone offers you a “definition of strategy”, pause. Do not debate it on taste. Do not be seduced by its neatness. Ask instead: what are the speaker’s strategic incentives? Are they selling you an artefact, a method, a cadence, a programme, a deck, a lens? Or are they trying to name the phenomenon itself?

If it is a product definition, it will quietly smuggle in a preferred tool and call it “strategy”. It will be partial by design. If it is a phenomenon definition, it will survive outside the boardroom. It will make predictions about where else strategy should be visible. It will draw a boundary that can fail. It will name the constraint set.

I am not asking you to “believe” my theory. I am asking you to take a wager that is testable in principle: if strategy is real, we should be able to recognise its functional invariants across species, systems, and domains. And if the definition cannot travel, perhaps it was never a definition at all.