Armenia had an alliance. Nagorno-Karabakh was outside the small print.

Nikol Pashinyan said the Russia-led CSTO had failed Armenia in 2021 and 2022, later freezing participation “in practical terms”. Armenia did not formally leave, and Russia’s base remained.

The dispute was about what the guarantee covered and when partners would act, not whether a treaty existed. Collective defence is most reassuring in peacetime, when nobody has yet asked its members to define the boundary.

The sales presentation will focus on performance, price and partnership. It is less likely to dwell on definitions, exemptions, territorial limits, notice procedures and the political interpretation of treaty language. Yet that is where the real leverage sits once presence, membership or a formal agreement is taken as proof that help will arrive in the required form. That is the control point hiding behind the headline fact: Armenia had an alliance.

For Armenia, this was not an accounting issue. The controlling party could honour the agreement in form while declining the specific action the threatened state expected. The contract might remain alive, the diplomatic relationship might remain cordial and the military effect would still be a missing capability. For Armenia, the warning is captured by what followed: Nagorno-Karabakh was outside the small print.

A serious review begins where the demonstration ends. What national capability covers the space outside the agreement? Which territory, actors and types of attack are actually covered? Is assistance automatic, consultative or subject to a fresh vote?

Resilience has to be purchased deliberately. The programme should translate treaty language into specific military planning assumptions, identify every exemption and grey zone before a crisis and retain national options outside the partner's mandate. If those items are unfunded, the independence claim is unfunded too. That exposure sits behind the claim that 'Armenia had an alliance'.

This is where scenario planning has to become specific. Do not write 'seek alternate supplier' in the risk register and move on. Name the supplier. Confirm the legal route. Price the integration. Hold the stock. Train the people. Until those steps are complete, the alternative is a comforting noun rather than an operational capability. The exercise has a concrete starting point in this case: Armenia had an alliance.

For Armenia, the small print becomes very large when shooting starts.

This should affect the score, not only the footnotes. Delivery credibility, substitution time, national repair depth and freedom of use deserve explicit weight in the evaluation. Otherwise the technically impressive bid wins on paper while the operationally usable bid is marked down for carrying the cost of being dependable. The headline fact, 'Armenia had an alliance', should carry that exposure in its score.

Legal precision is not cynicism about alliances. It is how military planners convert a political promise into forces, timings and responsibilities. That distinction matters here: Nagorno-Karabakh was outside the small print.

The force commander needs a different set of answers. Who decides whether an exemption applies? What notice or approval process delays action? Can commercial supply continue when government aid stops? Procurement success is not contract signature. It is usable strength on the required date.

The final test for Armenia is simple: what exact military action follows when the crisis occurs at the least convenient edge of the agreement? If the honest answer begins with 'provided that somebody else agrees', that condition belongs on the first page of the procurement decision, not in an annex nobody reads. The first half of this headline, 'Armenia had an alliance', is where that test begins.

A legal distinction that looks narrow in a conference room can become the exact place where protection ends during a crisis. The small print becomes very large when the shooting starts. That exposure sits behind the claim that 'Armenia had an alliance'.