The dances consist of about 30 seconds of movement that you repeat over and over, maybe fifteen to twenty times. Each time through you and your partner are dancing with a new "neighbor" couple, but you're making the same motions as the last time. Since it's so repetitive, by the end of the dance everyone generally has learned the dance fully, dancing from memory, and the caller will stop calling or "drop out".

Calls typically have four, two, and one beat variants. For example, you could prompt "long lines forward and back" (four beats), "long lines" (two), or "lines" (one). The more beats you use for the call the earlier your call is starting and so the more warning the dancers have about what they're going to do next. In dropping out you slowly reduce how many beats you're using to call, all the way down to eventually not giving any call at all (zero beats).

As I was learning to drop out, one thing that became clear was that you want to help the dancers just as much as they need, and no more. If you don't give enough help, things don't go well, as you might expect, with people getting confused about what to do next and mixing up parts of the dance. But if you give too much help you run into the opposite problem, which is that they start to rely on your calls. The goal is for people to end up with the dance in their heads and their bodies, and to rely on non-verbal cues from the other dancers to support their memories. Calling too long slows down this process.

In the moments before prompting each figure I would be making unconscious predictions about how many dancers would remember what came next. Factors would include how many times through the dance we'd been, the general skill level of the hall, how obviously this figure flowed from the previous one, whether this part of the dance could be confused for a different part of the dance, and how many people seemed to remember last time through. The more I thought people had learned the dance, the shorter a call I would give, all the way down to no call if I thought they wouldn't need one.

Sometimes I would judge incorrectly, and not give enough information, and I could see that it was a little rough, maybe slightly breaking down before people recovered. Other times I would give too much information, and in those cases I wouldn't get much feedback that I had judged poorly, just a weak sense that the hall seemed more solid than I had expected and maybe I had prompted a bit more than I could have.

Calling different nights with different approaches, it was very clear that the more careful I was to avoid giving calls that weren't actually needed, the faster people learned. Not only would I be able to drop out sooner that dance, but the following dances would go better because newcomers were figuring out how to memorize dances and how to follow the people around them.

Giving enough support but no more than needed is actually a really general skill, and one I've found useful in a lot of areas. Recently I've noticed how much it comes up in parenting. In so many areas kids move from needing you to do something for them, to being able to do it with some help, to being able to do it fully on their own, and you want to provide just the right amount of help throughout.

For example, a sufficiently small child can't dress themself at all, and you just put their clothes on them while they wriggle uncooperatively. Then their wriggles start to become cooperative, and you work with them to get their legs into their pants. Then they start getting to the point where there are pieces they can do unassisted, and you make space for them to pull their pants up once you've gotten their feet started. Later they get to where they can often do it by themself, but might need help if something is unusual (a pant leg starts inside out) or something goes wrong (both legs in the same hole). And eventually they don't need assistance at all.

I try to notice how much help my kids need, let them figure things out by themselves, and leave space for them to ask for help if they want it. And I feel like learning to call was pretty good practice!