The distinction matters later

Let it be recorded that the heart, though smaller than a fist, has ruined armies more efficiently than winter, and that an alliance is tested most severely not when its members face a common enemy, but when one of them must decide what the other is owed.

The difficult question is not whether two rulers can make a promise. They can. It is whether the promise still has a shape when the danger that called it forth has passed. A neighbor who sends soldiers during a siege may be a friend, a creditor, or a person who has made a very handsome calculation. Often he is all three, and the troops do not arrive carrying separate banners for each motive.

The distinction matters later. If the threatened ruler asks for help again, does he appeal to the old promise, or must he offer something new? If he offers a province, a marriage, or the right to cross his lands, he may discover that he has not renewed an alliance so much as taught his ally the price of his need. Gratitude is a poor ledger. It records the gift with great tenderness and the debt with alarming precision.[1]

This is where peaceful years become dangerous. A court may see no battle and conclude that its partners are reliable; meanwhile, each partner has been keeping a different account. One remembers the winter when its grain was shared. Another remembers that the wagons arrived after the first thaw, when the hungry had already begun to leave. Both accounts may be true. They are not the same relationship.

Even Lady Maelin, whose judgment I would trust above that of most councils and several star tables, could not make such a difference disappear merely by smiling at the right moment. A marriage between houses can make an old promise harder to break, but it can also give every slight a private address. When a border lord is refused aid, he may blame the treaty. His sister, married into the other court, may blame the husband who did not speak for her. The public grievance travels by road; the private one takes the stairs.

So the strongest alliance is not necessarily the one with the grandest vow. It may be the one whose members can disagree without making each disagreement a trial of affection. A useful ruler asks not only, “Will you come when I call?” but also, “What will you think I have promised by asking?” That second question is less flattering. It is also less likely to leave a kingdom astonished by the bill.

The small signs should be kept. Who sends a messenger before the harvest is counted? Who returns a borrowed banner clean, and who returns it with a new crest sewn over the old? These are not proofs of loyalty, exactly; proof is a word people use when they wish to stop looking. They are evidence of how carefully one party believes the other deserves to be remembered.

And memory, in politics as in love, is not a storehouse. It is an argument conducted over years, with each side choosing which detail to place in the margin. The alliance survives not because the two rulers recall the same past, but because they can still make a future together without pretending that they do.[2]