Intimacy II

Further to my thoughts this morning on the separation of public and intimate relationships, it occured to me that I missed some interesting connections.
I wrote that we need emotional commitment where we can't achieve commitment via public enforcement (contracts) because the considerations required can't be specified precisely enough (perhaps because flexibility is itself a key consideration). Possibly more important is the fact we can't enforce publicly (using the law) something that is supposed to happen in private, without witnesses. This came up before when I defended old-fashioned courtship patterns as a way of avoiding the unpleasantness that can result from being alone without witnesses with an untrusted partner.
The concept that keeps coming up is the cost or difficulty of enforcing any arrangement. Whether I am talking about intimate relationships, the basis of property, the structure of government, law and order, or the business models of entertainment products, it keeps coming up as the decisive factor. Either I have a bee in my bonnet about it, or it is being generally overlooked: treated as a minor implementation detail to be worked out later. Or both, I suppose.

Another stray thought on drawing a boundary around the intimate is Linus Torvalds' famous quote: Software is like sex -- it's better when it's free. Taking the idea altogether too seriously, what might there be about the writing of software that makes it more suitable to being motivated by emotional commitment rather than public bargain?
It might just be the undefinability of the requirements. A piece of software isn't much to look at, it's very difficult to assess its value in advance. Even if you can determine that it functions correctly, that's not a complete assessment -- quality of software is notoriously difficult to define. If you have the freedom to take what you need from software, that is perhaps more valuable than a predefined functional specification.