How a war is built

1/4 · ANALYSIS · 28 July 2026

Thirty years to a war

The context

Wars rarely erupt from a single cause. The one between Russia and Ukraine, in its fifth year of full-scale fighting by 2026, has roots reaching thirty years before the first shot — a useful case for understanding how an escalation is built, often with neither side fully able to stop it in time.

The roots (1990-2008)

In 1990, during talks on German reunification, US Secretary of State James Baker told Gorbachev that NATO jurisdiction would not move “one inch eastward.” Declassified National Security Archive documents (2017) show similar assurances repeated by other Western leaders in the following months. Historians disagree on how narrowly these applied to Germany alone; they agree that **none was ever formalised in a treaty**. In 1997 the NATO-Russia Founding Act, with Russia’s signature, affirms every state’s right to choose its own alliances. Two rounds of enlargement follow (1999, 2004). In 2008 the Bucharest Summit declares that Ukraine and Georgia “will become members,” with no timeline — an ambiguity that would prove more destabilising than a clear yes or no.

The interference never fully reckoned with (2014)

In February 2014 a phone call between US Assistant Secretary of State Victoria Nuland and US Ambassador Geoffrey Pyatt was intercepted and posted online — the US never disputed its authenticity. Reported by the BBC, it shows Nuland openly discussing who should lead Ukraine’s future government: “I think Yats is the guy [...] Klitsch and Tyahnybok need to stay on the outside.” After Yanukovych fled, Yatsenyuk did become prime minister. Russia’s annexation of Crimea and the Donbas war followed. The Minsk agreements (2014-2015) remained largely unimplemented over an unresolved ambiguity about which condition should be met first. A 2022 admission by former Chancellor Angela Merkel — that the agreements also served to “give Ukraine time to get stronger” — fed Russian distrust, though it remains an ex-post political statement, not an independently verifiable fact.

The ideological leap (2021)

In July 2021 Putin published the essay “On the Historical Unity of Russians and Ukrainians,” later added to mandatory reading for the Russian military. It argues Russians and Ukrainians are “one people” and questions the legitimacy of Ukraine’s borders, calling Ukraine an “anti-Russia project.” From here on the problem is no longer only military security, but Ukrainian nationhood itself — something no technical deal on missile bases could resolve. December 2021 brought a formal Russian ultimatum, rejected by the West.

The invasion and the miscalculation (2022)

On 24 February 2022 Russia launched its full-scale invasion. Russian planners expected capitulation within days. Per Foreign Affairs (Charap and Radchenko, based on direct interviews with negotiators), Russia’s withdrawal from Kyiv in late March 2022 — presented by Moscow as a goodwill gesture — was in fact a forced retreat, after “the Russians had overestimated their capabilities and underestimated Ukrainian resistance.”

The window that closed (Istanbul, April 2022)

By mid-April 2022 Russia and Ukraine had reached a nearly complete draft treaty: Ukrainian neutrality in exchange for security guarantees from the five permanent UN Security Council members. A Ukrainian negotiator publicly stated “we were very close to ending the war.” Why the deal collapsed remains contested: Moscow blames British pressure (never independently confirmed); the prevailing Western account attributes it to Kyiv’s choice, emboldened on the battlefield and shaken by the Bucha revelations, to press on rather than accept painful concessions.

The opposite effect

The December 2021 ultimatum and the invasion produced the exact opposite of Moscow’s stated goal: not a NATO retreat, but Finland (2023) and Sweden (2024) joining, doubling the Alliance’s direct border with Russia.

The symmetry test

The legal and behavioural planes must be kept separate. Legally, the Russian invasion does not meet Article 51 requirements: there was no Ukrainian armed attack to repel, and this remains true regardless of what follows. Behaviourally, none of the three main actors comes out clean: the 2014 US interference, the Minsk ambiguity attributable to more than one party, Ukraine’s choice to fight on after Istanbul. None of this rebalances the legal plane.

Editorial judgment

The Russia-Ukraine case shows a pattern that will recur, with variations, in other conflicts: security dilemma, a narrative denying the other’s legitimacy, diplomacy left deliberately ambiguous, miscalculation, and at least one real diplomatic window left to lapse. Understanding this mechanism is not about distributing blame equally — the violation of international law remains a distinct fact — but about understanding why wars, once triggered, are so hard to close.

Fonti: BBC, Nuland-Pyatt call transcript (7 February 2014) · Christian Science Monitor (6 February 2014) · National Security Archive, George Washington University — "NATO Expansion: What Gorbachev Heard" (2017) · Atlantic Council / OSW Centre for Eastern Studies — Putin’s essay (July 2021) · Foreign Affairs, Charap & Radchenko — "The Talks That Could Have Ended the War in Ukraine" · European Council on Foreign Relations — "Ukraine, Russia, and the Minsk agreements: A post-mortem"

2/4 · ANALYSIS · 28 July 2026

A deal that worked, torn up anyway

The context

Sometimes a conflict isn’t born from a lack of diplomacy, but from its unilateral interruption — even when it was working. The path that led the United States and Israel to bomb Iran directly twice in under a year, killing its Supreme Leader, begins with a deal the international watchdog certified as being respected.

The deal (2015-2018)

In July 2015 Iran, the United States, Russia, China, the UK, France and Germany signed the Joint Comprehensive Plan of Action (JCPOA): strict limits on Iranian uranium enrichment in exchange for lifting sanctions, under an International Atomic Energy Agency (IAEA) inspection regime Washington itself called “unprecedented.” In 2017, all four IAEA quarterly reports certified full Iranian compliance. On 8 May 2018 President Trump unilaterally withdrew the US from the deal, publicly calling it “a horrible, one-sided deal.” A detail often omitted but verifiable: the White House withdrawal memorandum identified no Iranian violation occurring during the Trump administration — only alleged 2016 violations, under Obama, concerning heavy-water stockpile limits. US sanctions were reimposed.

The predictable consequence (2018-2024)

Stripped of the deal’s benefits but no longer bound by its limits, Iran progressively rolled back its commitments: increasing enrichment, restricting inspector access. Breakout time — how long it would take to accumulate enough material for a device — fell from over a year to 3-4 months, per Center for Arms Control and Non-Proliferation estimates. In 2020 a US drone killed General Qassem Soleimani in Baghdad. After 7 October 2023 and the Gaza war, Iran and Israel exchanged direct strikes twice in 2024.

The 12-Day War (June 2025)

On 13 June 2025 Israel launched surprise strikes on Iranian nuclear and military sites, killing senior Revolutionary Guard commanders. On 22 June the United States struck Fordow, Natanz and Isfahan directly with bunker-buster bombs — Trump called it a “limited, one-off” mission. A ceasefire followed on 24 June. Estimates of the actual damage diverge sharply by source: the US says “destruction”; a leaked US intelligence report and Israeli intelligence describe serious but not conclusive damage, with the program delayed by months, not years.

The failed talks and the second war (2026)

In February 2026 indirect Oman-mediated talks on a new deal failed — Iran’s foreign minister signalled willingness to make concessions, but Trump said he was “not thrilled” with the outcome. On 28 February 2026 the US and Israel launched a new large-scale offensive — nearly 900 strikes in 12 hours — this time explicitly declaring the goal of regime change. Supreme Leader Ali Khamenei was killed in the opening strike; his son succeeded him. A conditional ceasefire, mediated by Pakistan, took effect on 8 April 2026.

The symmetry test

This is where the standard applied elsewhere in this series becomes harder to apply, not easier. If the yardstick for judging the Russian invasion of Ukraine is “no ongoing armed attack to repel = not legitimate self-defence under Article 51,” the same yardstick applied to the 2025 and 2026 US/Israeli strikes — against a nuclear program not yet weaponised, with no ongoing Iranian armed attack — leads to the same legal conclusion. The doctrine of “anticipatory self-defence” against a future threat is not settled international law. Killing a head of state and explicitly declaring a regime-change objective are, as a matter of principle, hard to distinguish from conduct the West condemns without hesitation when attributed to other actors.

Editorial judgment

What links this case to the previous one in the series is not a symmetry of blame — the facts and violations remain distinct — but the mechanism: a working deal, unilaterally broken by one of its own parties; a self-feeding escalation because each side reads the other’s move as an existential threat; and a diplomatic window (the Oman talks of February 2026) left to lapse shortly before armed force took its place.

Fonti: Britannica — "12-Day War" (June 2025) · Britannica — "2026 Iran war" · House of Commons Library — "Israel/US-Iran conflict 2026" · CFR, Global Conflict Tracker — "Iran's War With Israel and the United States" · Congressional Research Service — "U.S. Conflict with Iran" (26 March 2026) · Center for Arms Control and Non-Proliferation — "The Iran Deal, Then and Now"

3/4 · ANALYSIS · 28 July 2026

Thirteen years to topple a man the West had already given up on toppling

The context

Few wars show as clearly how a conflict can drag on by sheer inertia even after all its main external backers have stopped believing in it. Syria from 2011 to 2024 is the textbook case.

From uprising to proxy war (2011-2015)

The 2011 Arab Spring protests against Bashar al-Assad turned into civil war after violent government repression. From 2013 the conflict rapidly internationalised: the CIA launched the secret Timber Sycamore program to arm “moderate” rebels — already covered in detail, including the failure of arms vetting and the documented links of some funded groups to jihadist networks, in The US armed the Syrian rebellion for five years. In 2015 Russia intervened directly on Assad’s side with its air force, while Iran and Hezbollah provided ground support — turning Syria into the decade’s principal proxy battleground.

Stalemate and the silent abandonment (2017-2024)

In 2017 the Trump administration shut down Timber Sycamore, in what a US official told the Washington Post was an explicit signal to Moscow — “Putin won in Syria” — days before the first Trump-Putin meeting at the G20. From here the conflict entered a seven-year stalemate: Assad reclaimed territory with Russian-Iranian support, but without political normalisation, while the US and Europe stopped actively investing in regime change, settling for sanctions.

The sudden fall (November-December 2024)

On 27 November 2024, Hay’at Tahrir al-Sham (HTS), led by Ahmed al-Sharaa, launched a surprise offensive in Aleppo province — a day after an Israel-Hezbollah ceasefire in Lebanon took effect, reducing available Iranian/Hezbollah military support for Assad. Aleppo, Syria’s second city, fell on 1 December; Hama on 5 December; Homs, the third city, on 7 December. On 28 November Assad personally flew to Moscow to request Russian military aid — the request was denied. He returned to Syria, lying to his own commanders that aid would arrive. On 8 December 2024 rebels entered Damascus meeting minimal resistance, as government forces abandoned their positions; Assad fled to Russia, where he was granted asylum. The end of thirteen years of civil war and fifty years of Assad family rule.

The aftermath, still unstable (2025-2026)

Al-Sharaa became interim president (January 2025); a transitional government was established in March 2025. Post-Assad Syria in 2026 is not a closed case: 15.6 million of Syria’s 23 million people need humanitarian assistance, and Turkey’s and Germany’s push for refugee returns continues despite UNHCR formally ruling out the possibility of safe return — already covered in detail in The UN says it isn’t safe to return to Syria.

Symmetry test and escalation mechanism

The Syrian case follows the same pattern as the other cases in this series, with one significant variant: here it is not a military miscalculation that produces the turning point, but the simultaneous exhaustion of resources of multiple external powers on other fronts — Russia tied down in Ukraine, Iran and Hezbollah weakened by confrontation with Israel — that creates the window for a sudden collapse. It’s a useful reminder for the other cases in this series: a frozen conflict is not a resolved conflict, and it can reopen not through new aggression, but simply because attention shifted elsewhere.

Editorial judgment

Russia’s refusal to help Assad — an ally on whom Moscow had spent years of direct military intervention — is perhaps the most telling fact in this case: it shows that no external support, however long-standing, is guaranteed once the costs elsewhere outweigh the strategic value of continuing to prop it up.

Fonti: CNN — "The Assad regime ruled Syria for 50 years. Here's how it fell in less than two weeks" (December 2024) · EUAA (EU Agency for Asylum) — "Country Focus: Syria" (March 2025) · The Media Line — "The 12-Day Revolution: How Syria's Regime Collapsed After Years of War" · Gulf News — "Syria rebels' lightning offensive: 11 days to the fall of Damascus"

4/4 · OPINION · 28 July 2026

How Europe could end up at war without choosing it

Methodological premise

This is not a news piece. It is a declared foresight exercise — what could happen, not what is happening — built on real, already-verified incidents, not pure speculation. Where forecast ends and fact begins is always explicitly flagged.

What is already happening

Between August 2024 and February 2026, per an International Institute for Strategic Studies report (July 2026), hundreds of drones flew over the airspace of a dozen NATO states plus Ireland — including bases tied to the Alliance’s nuclear deterrence, such as Ramstein in Germany. Poland invoked NATO Article 4 in September 2025 after 19 Russian drones entered its airspace. Romania suffered, in May 2026, a drone strike on a residential building in Galați, injuring two people — called by the Romanian president "the most serious security incident" since the war began. In Latvia, the defence minister and prime minister resigned in June 2026 after drones (likely Ukrainian, diverted by Russia) violated national airspace without timely public alerts. On the Iran front, during the February-April 2026 war, British bases in Bahrain, Qatar and Cyprus were directly hit by Iranian retaliation, with the RAF deployed defensively — direct European military involvement that has already happened, not a hypothesis.

Scenarios — Russia

The most likely is not a deliberate attack but the incident that spirals out of control: a drone — Russian or Ukrainian, its origin often ambiguous for days — hits a target with multiple casualties in a NATO country before warning systems work. With each subsequent incident, the threshold for a collective response (Article 5, not just Article 4) lowers through statistical accumulation, not a single political decision. Other, less likely but non-negligible scenarios: a formal judicial attribution of sabotage against undersea infrastructure (on the Nord Stream model, already covered elsewhere on this site); an air or naval engagement error between NATO and Russian fighters; a gradual escalation if European long-range weapons strike ever deeper into Russian territory, pushing Moscow to explicitly strike the supplying country’s own territory, not just Ukraine’s.

Scenarios — Iran

The precedent that has already occurred (British bases hit, RAF deployed) could repeat itself with a European power less resolute about staying "purely defensive." A disruption of merchant shipping in the Strait of Hormuz involving European vessels would turn an energy crisis into an incident requiring a direct naval response — a pattern already seen in the 1984-88 "Tanker War." More speculative, but not to be ruled out: an attack on European critical infrastructure attributed to Iranian proxies.

What diplomacy should do — the Russia/NATO front

The instrument that existed and is now missing is the Vienna Document on confidence- and security-building measures (OSCE, 1990-2011): annual military information exchanges, prior notification of exercises, mutual inspections among 57 states. Russia suspended it de facto from January 2022, formally from March 2022. This is the structural gap behind every drone incident: without prior notification or shared verification, any unidentified object remains ambiguous long enough to become dangerous. Concrete measures available: a bilateral technical deconfliction channel, separate from political channels, modelled on the 1972 US-USSR incidents-at-sea agreement or the US-Russia deconfliction line over Syria (2015-2024, which functioned even amid total political hostility); a rapid-attribution mechanism under OSCE auspices, which in 2026 has Switzerland holding the rotating chair; symmetrical pressure on Kyiv too, to limit long-range drone launches near NATO borders. None of these technical measures, however, replaces a negotiation on the underlying conflict: it reduces the frequency of incidents, not eliminates them.

What diplomacy should do — the Iran/Europe front

EMASoH (European Maritime Awareness in the Strait of Hormuz), a French-led mission launched in 2020 based in Abu Dhabi — a diplomatic track plus a military one (Operation Agenor), 9 European countries — already exists, explicitly designed for de-escalation through joint diplomatic and military presence. Recent sources (March 2026) describe it, however, as under growing pressure relative to its original mandate. Concrete measures: strengthening EMASoH’s diplomatic track, today less visible than the military one but specifically designed for regional dialogue; reinstating the E3 (France, Germany, UK) as a recognised negotiating channel on the nuclear file, not merely an observer of the US-Iran bilateral track; explicit European non-belligerence messaging, to formally distinguish the EU’s position from that of the US/Israel. Here too, the underlying limit remains the same: no European naval or diplomatic measure resolves the root cause, which is the 2018 collapse of the JCPOA.

Editorial judgment

On neither front does Europe today have the tools it had before 2022, or the ones it was conceived with in 2020. The most honest question is not "what should diplomacy do" in the abstract, but whether Europe is investing to rebuild dismantled channels or merely managing symptoms with tools undersized for the current scale of risk — the latter appears to be the reading most consistent with the facts gathered.

Fonti: IISS — drone incursions report (July 2026) · Al Jazeera — Romania drone incident (May 2026) · NATO — Article 4 statement (September 2025) · CS Monitor — Latvia government crisis (June 2026) · House of Commons Library — UK bases hit in Iran (2026) · Auswärtiges Amt / Arms Control Association — Vienna Document · European Leadership Network / GRC — EMASoH