Retaliation as Architecture

I. When Threats No Longer Hold

The European security debate thinks in twentieth-century categories: more tanks, more brigades, higher spending, sharper threats. This answer presupposes three things that no longer hold. First, a reliable American security guarantee. Second, democratic societies willing to escalate against an aggressor. Third, opponents who calculate rationally rather than playing the strategy of the unpredictable.

In Europe the American distancing is still read as the mood of a single president. But it has three layers that reinforce one another. Economically, the old division of labour is over, under which America supplied security and Europe developed its civilian economy beneath that umbrella. The White House’s 2025 National Security Strategy names energy, arms sales and cloud infrastructure under American law explicitly as instruments of power. Security guarantees are contracts with consideration clauses; what is new is only how openly they are stated. Religiously, a political theology has entered the Pentagon that is barely understood in Europe. Pete Hegseth, Secretary of Defense since 2025 and bearing the title Secretary of War since September 2025, attends a congregation of a church network around the pastor Doug Wilson which, according to scholars of religion, is strongly shaped by Christian Reconstructionism, a current that seeks to establish a legal order founded on the Bible. In an apparatus hitherto built on dutiful compliance, it is no longer the hesitant who rise, but the certain. Ideologically, a current has gained power in the Trump administration that regards liberal democracy not as a heritage but as an obstacle. We described it in “The New Right”.

Whoever builds on a predictable American policy is hoping that one election will clear away all three layers at once. That hope is unfounded. The usual answer is rearmament. In May 2025 Friedrich Merz announced that the Bundeswehr should become the strongest conventional army in Europe, and repeated this in Munich in February 2026. The frame of reference is telling: Europe is the yardstick by which German strength is measured, not the body that builds it. And even a stronger army would not change the fact that a democracy remains exposed to nuclear blackmail as long as the blackmail works as a threat.

II. What Russia Wants

Two readings face each other in Europe: Russia wants to conquer Europe, or Russia only wants its security zone and can otherwise be appeased. Both miss Russian strategy. It counts neither on a leap westward nor on renouncing the West. It counts on waiting.

Russia is the largest country on earth and still believes it needs more land. This has two roots. The first is the North European Plain, a surface from the Atlantic to the Urals over which armies move as over a table. They have moved: the Poles in 1610, the Swedes in 1708, the French in 1812, the Germans in 1918 and 1941. The Russian answer was always the same: depth of space as a substitute for defence. Catherine the Great is credited with saying she could defend her borders only by extending them. Whether she said it is not documented. That it describes Russian practice is.

The second root lies in the east. For almost two and a half centuries the Russian principalities paid tribute to the Golden Horde. Many historians see in this one of the roots of centralised Muscovite autocracy; how much weight it carried beside the Byzantine heritage is disputed. Undisputed is that Russia has since been governed without rule of law, secure property or an independent middle class, and that a regime which may not admit this weakness attributes it to external enemies.

From this follows what Ukraine is for this regime. It is not one piece of land among others. It comes from the same Soviet inheritance and is developing, with all its difficulties, in a different direction. A successful Ukraine refutes the Russian narrative that the system is inevitable. That makes it more dangerous to the Kremlin than any NATO enlargement. For Europe this means: the direction of Russian policy is more predictable than it seems. What is not predictable is the price the regime is prepared to pay for that direction.

III. The Principle: Taking the Threat Out of Human Hands

In game theory, the simplest successful strategy is called “tit for tat”: one cooperates first and then answers each action of the other with the same. The problem in politics is not the strategy but the moment of its execution. Whoever is to answer an aggression with a counter-action needs the will to escalate in the moment of crisis, and that is exactly what democracies lack against an opponent who threatens with unpredictability.

The eight concepts of this essay therefore follow one principle: the threat is taken out of the hands of the politician who may be weak at the decisive moment and placed in a structure that is less weak than a single person under pressure. The counter-reaction is not decided in the moment of crisis but built beforehand, and everyone involved knows it in advance.

Tit for tat has a known weakness: misunderstandings can trigger a spiral in which each side believes it is only responding. In autumn 1986 the United States and the Soviet Union expelled each other’s diplomats in several rounds, until some 260 Soviet local employees had to leave the American embassy in Moscow and the ambassador drove himself to work. Neither side wanted to escalate; both did. The architecture meets this weakness with unambiguous triggers. A nuclear detonation is detectable seismically and radiologically. An attack on the Suwałki Gap is visible as troop movement, with time, place and perpetrator.

Honesty requires distinguishing how far the shift reaches in each case. Strongest is physical self-enforcement: in space, the debris cloud punishes the attacker without anyone having to decide anything. In the middle stands documentary enforcement: a register builds itself, but ports, banks and authorities must carry out the consequences. Weakest is contractual enforcement: between trigger and consequence stands a state that must act when it counts. A treaty is not a saw with a finger guard. But it can prepare the decision so far that failing to act becomes the breach of a public promise, and thus more costly than acting.

IV. Eight Concepts

NUET, outlawing the use of nuclear weapons. Whoever uses a nuclear weapon against another country is excluded from the world economy for at least fifty years, without a new vote, without a veto, without a committee; so is whoever helps. The architecture holds if states representing about two thirds of the world economy sign it. Its natural carriers are not the United States but Europe, China, India, Japan, Brazil and the Global South. It works against every nuclear power that relies on blackmail. The exclusion must be carried out by the signatory states themselves; this is where the architecture remains most vulnerable.

RIEGEL, geographic reciprocity. Since Finland and Sweden joined NATO, the Baltic has been an inland sea of the alliance. A Russian attack on the Suwałki Gap between Poland and Lithuania would, as agreed in advance, trigger the sealing off of Kaliningrad and the closure of the Gulf of Finland. The exclave, with around a million inhabitants, would be cut off, St Petersburg trapped in the funnel. The attack would lose its point. The precondition is the unity of the littoral states.

MESH, resilience through decentralisation. Critical infrastructure is protected not by more guards at more cables but by a design in which destroying one node destroys only that node: local energy cells in each district, redundant networks, local reserves. The arson attack on the power supply in south-western Berlin in January 2026, which left tens of thousands of households without electricity for days, showed how vulnerable the centralised structure is. Finland and Switzerland are the models.

SHADOW, reversing the burden of proof at sea. Proving that a ship deliberately damaged an undersea cable is almost never possible in international waters. So sailing through a cable protection zone without a working transponder becomes an offence in itself. The ship is publicly registered, and the ports of participating states refuse it service. This hits the Russian shadow fleet, by estimates several hundred tankers, without a battle on the high seas.

GRADUS, a graduated response to covert attacks. Russia’s hybrid strategy lives by staying, with many small acts, below the threshold at which a political response would be due. GRADUS keeps a public register of all documented acts. Three thresholds fixed in advance trigger responses fixed in advance. Every act counts, every threshold is known, and at the moment of triggering no unanimity is needed any more.

COSMOS, space as a common good. Here self-punishment need not be constructed; physics has already installed it. Whoever shoots down a satellite creates debris that also threatens his own satellites. COSMOS adds a liability convention, similar to liability for oil pollution in maritime law, and at the same time regulates the strategic role of private networks such as Starlink.

DEMOS, protecting the formation of opinion. In December 2024 Romania’s Constitutional Court annulled the first round of the presidential election. The problem was not a forgery but an algorithm. DEMOS relies on transparency rather than censorship: labelling of artificially generated political content, disclosed recommendation logic during elections, disclosed funding. It is the only concept that relies more on immunisation than on reciprocity.

POLIS, the inner pillar. The greatest danger to democracies currently comes not from outside but from movements waiting in every constituency for the moment the established structures fail. POLIS answers with four elements: individual decisions can be delegated to competent representatives without handing over the whole mandate; qualifications for offices are publicly verifiable; decisions are taken at the level where their consequences fall; no one holds the same office for more than two terms. A democracy that itself offers clarity and speed does not leave that need to those who would satisfy it in authoritarian ways.

V. Leniency as a Component

A principle that places threats in structures has a sister: the principle of placing leniency in structures. In “The Construction of Mercy” we showed how mercy got into the machine, not as an attitude but as a bill and a rule. The same movement, from the moment of human decision into the construction built beforehand, also carries this architecture. It would be incomplete if it constructed only the harshness and not its limit as well.

Game theory knows this limit as generous tit for tat: whoever answers every breach but occasionally forgives one breaks the spiral of misunderstandings without becoming exploitable. For the architecture this yields a distinction. Where the trigger is unambiguous, a nuclear detonation or an attack on the Suwałki Gap, there is no leniency; every exception would hollow out deterrence. Where it is ambiguous, with covert acts and shadow ships, leniency belongs in the construction. Entries in the GRADUS register expire after a fixed period if no new ones are added. Whoever demonstrably cooperates moves down a level. A ship that switches its transponder back on and allows an inspection is removed from the SHADOW list. The harshness remains predictable, and so does the way back.

VI. A Carrier: the Navy in Gdańsk

Architectures need someone to enforce them. Since the failure of the European Defence Community in 1954, a European army has been proposed again and again and dropped again and again. Friedrich Merz, too, did not speak out for one in Munich in 2026, but for a “self-sustaining European pillar” within NATO.

We propose a smaller entry: a European defence union that begins with a single service, the navy, and a single headquarters, in Gdańsk. The navy is the service with the weakest national ties; fleets have sailed in combined formations for centuries, and there are already standing NATO naval groups in the Baltic, the North Sea and the Mediterranean. What is missing is not substance but common command. Gdańsk lies at the focal point: Kaliningrad, the Gulf of Finland and the Suwałki Gap are within reach. Poland is the state with the most determined stance towards Russia. And the city carries a double history: here the Second World War began in 1939 with the attack on the Westerplatte, here Solidarność began in 1980 at the shipyard. A headquarters in Gdańsk would be neither in Berlin nor in Paris; no one could read it as a bid for dominance. The other services could follow once the navy has proved itself.

VII. Arming Wisely Instead of Filling Wish Lists

Together, Europe spends several times what Russia spends on defence, yet achieves only a fraction of the effect. Every country arms according to its own doctrine, its own standards, its own supply chains. Arming wisely would mean: first an honest inventory, then an analysis of the likely threats, from that a list of gaps, and only then orders. Not spending less, but differently.

Four gaps are visible already. Defence against cheap drones, which in Ukraine knock out expensive weapon systems; Europe still mostly buys the expensive ones. Sovereignty in space: Galileo is running, IRIS² is ordered, but there is no European alternative to Starlink. Mass production: the trade press estimates annual capacity for the Taurus cruise missile at forty to sixty units, a few per month. And the protection of cables and networks, for which there is no industrial base, national or European.

VIII. Why China Would Have to Lead

A Western initiative that seeks Chinese participation afterwards is perceived as an instrument of containment and fails. An initiative from Beijing, joined by Europe, India, Japan and Brazil, would have different prospects. Three reasons speak for this. NUET works against every nuclear power that relies on blackmail, the United States included; a Chinese leadership that has lived through two Trump administrations knows the value of such a rule. The principle of turning the attacker’s force against himself is East Asian heritage, from aikido to Sun Tzu. And China has its own interest in clear rules in space and in an order that no single actor can destabilise through blackmail. In May 2026 President Trump travelled to Beijing with seventeen chief executives. Whether Beijing would take on such a role, we do not know. That is for sinologists and diplomats to judge.

IX. How This Essay Could Be Refuted

Unlimited tolerance for costs. All concepts presuppose an opponent who weighs costs against benefits. In Ukraine Putin accepts losses that exceed every Western calculation. Against such an actor the concepts shift the calculation, but they do not decide it. The higher the tolerance for costs, the weaker the lever.

Preparation and enforcement. RIEGEL, GRADUS and SHADOW presuppose that the EU and NATO agree in advance; with Hungary and Slovakia that is hard. And even where it succeeds, the moment of crisis returns at enforcement. Germany’s accident insurance of 1884 worked because trade associations inspected the businesses and a state stood behind them. No such authority exists between states. The architecture can make a failure to enforce costly, but not impossible.

Humanitarian harshness. NUET deliberately knows no humanitarian exceptions, because exceptions would hollow out the concept. That is consistent and morally hard. Whether democracies would hold to a rule whose consequences the aggressor’s population would bear is an open question.

The missing test. None of the concepts has been legally examined or militarily war-gamed. The essay would be refuted if experts could show that the unambiguous triggers are not unambiguous after all when it counts, or that counter-reactions built in advance do not bind democracies more firmly than a decision in the moment of crisis.

X. Victory Without Battle

These concepts are not a doctrine but answers to a simple question: what can an architecture achieve when threats no longer hold? Some answers may be naive. The question is not. Asking it seems to us more urgent than the next speech about defence spending.

If we have learned to teach a machine whom it must spare: why should we not be able to teach a peace order whom it will inevitably strike?