On writing square dance material

In the past few days I called at two square dances. One was for Fall Challenge Weekend, where I called a tip of what I called “CX”, which is the C1 list plus a few calls and concepts from higher levels. The other was for Cambridge Challenge Series, which was a whole night; I called eight tips of alternating C1 and C2. Because of all that, I’ve been thinking quite a bit about square dancing, and particularly about writing material.

Units of calling

I’ve talked about the compositionality of square dancing in the past, where I analyzed the structure of individual calls. But calls fit into higher-level structures as well, and when calling, it’s often more convenient to think in these bigger structures. From smallest to largest:

Words. I’d say maybe half of all calls have more than one word in their name. Calls don’t make sense at the word level, but I’ve been to some improvised gimmick tips where two callers took turns saying calls; most recently in IAGSDC 2024. While this is mostly done at the call level, sometimes it’s at the level of words.

For example, one caller can say the A2 call “Swing”, and the other caller can say the word “Thru”, to instead form the Mainstream call Swing Thru. As another example, one caller can start with Switch to a, and the other caller can choose between Diamond or Hourglass, both being different A2 calls. A different kind of example is starting with $\frac{1}{2}$, which is a concept in itself, but not a full call.

I don’t think calling at the level of words is something people do outside these gimmick tips. However, the square dance program SD lets you type in a concept, and you can then type pick random call to let SD pick the completion for you. Of course, it’s a computer program, so the results aren’t always the most inspiring:

 3B^   2GV   3G^   4BV

 2B^   1GV   4G^   1BV

 1:
(waves)--> 1/2 ⏎

 1:   1/2,
--> pick random call ⏎

Pick Random Call: searching ...

 1:   1/2, swing the fractions 4/5

 3B>   3G>

 2B<   4G<

 2G>   4B>

 1G<   1B<

Pick Random Call: 1 out of 1

 3B^   2GV   3G^   4BV

 2B^   1GV   4G^   1BV

 1:
(waves)--> 1/2 ⏎

 1:   1/2,
--> pick random call ⏎

Pick Random Call: searching ...

 1:   1/2, swing the fractions 4/5

 3B>   3G>

 2B<   4G<

 2G>   4B>

 1G<   1B<

Pick Random Call: 1 out of 1

I’ve found it useful for getting inspiration when I want to use a particular concept.

Calls. A prized caller skill is the ability to sight call, which is improvising a sequence of calls. There’s several considerations to stringing calls together that’s been covered at length in other places; I like Coop Bellini’s rules of flow and Rich Reel’s calling notes. The most important consideration when sight calling, though, is resolution.

To resolve a square is to get the dancers back to where they started; a properly formed sequence starts from a squared set and ends at the same spots. There are many techniques for improvised resolution; take one look at the Callerlab knowledgebase. The technique I’m most familiar with is the double pass thru method, but the system I admire the most is the mental image system.

If resolution alone wasn’t difficult enough, you need to resolve the square while keeping other choreographic considerations in mind, like flow and hand alternation. And you need to come up with calls fast enough, so that you don’t keep the dancers standing around. And you don’t want to do the same things over and over again either. I have a deep respect for the callers who do improvise well.

Improvising calls is a task I’m not at all good at, but I’ve been meaning to learn a little bit. It’s useful when you mess up a sequence and need to get resolve without resorting to an Explode and Go Home. I suppose it’s also useful if you want the highest ratio of calling to preparation time.

Modules. In between fully improvising and preparing sequences in advance is memorizing modules, which take one formation to another. A sequence is the special case of a module that takes a squared set to a resolve, but there are more granular modules as well. The module endpoints are commonly the zero arrangements, per Callerlab’s naming scheme; particularly common are the zero line and zero box formations. Another common kind of module is a zero, which does nothing to a formation, though note that there are several kinds of zeros.

Sequences. As a Challenge caller, working at the level of calls or modules doesn’t quite fit my calling needs. In particular, my mind is not fast enough to improvise the kinds of things I like calling. So I usually write sequences in advance and call them live.

How long does it take me to write a sequence? It varies based on the type of sequence. It takes around a minute to call a sequence, and at my fastest, I can write a sequence in around a minute. For complex sequences, it can take up to thirty minutes; I’d say my median is around five minutes? It’s hard to tell, as I don’t try to be efficient with my sequence writing.

Calling via sequences, as opposed to sight calling, is also called calling using cards. The name comes from the traditional way that callers organize sequences, which is by writing them on index cards and sorting them into piles. I’ve also seen callers print sequences onto letter paper, one sequence per sheet. Beyond organizing sequences by level (you wouldn’t want to call a C4 sequence to a Plus square), the most common scheme seems to be organizing by difficulty. It’s helpful for following a sequence with easier sequences if the dancers struggled with the last one, or following with harder sequences if the dancers are succeeding.

The other option is to organize the sequences digitally. The most popular software for this is probably CSDS, but there are other options. Andy Tockman wrote sdb, a set of Ruby scripts, and webapp they wrote called sdj. I use a webapp I wrote called sdcard. The name refers to how it takes SD files as input; it’s also a nod to card calling and a pun. I’ve been meaning to make sdcard better, but work’s been paused due to my other squares-related code work (which I’ve also been meaning to write about).

Tips. Sequences are themselves grouped into tips. A tip is the span of time from when people get up and form squares to when a break is called, and it’s usually around twelve minutes long. There are two common structures for a tip: patter then singing call, and patter-only.

In the patter, the caller calls through a series of sequences; in a singing call, the caller calls over the instrumental track of a song in verse–chorus form. Singing calls take around three minutes, so tips that are patter plus singer have maybe eight minutes of patter. At Challenge, the vast majority of tips are patter-only. This usually means ten to twelve sequences, over around eleven minutes, with potential variance due to sequence length or redoing sequences.

My understanding is that, while all Challenge callers prepare sequences in advance, it’s less common for a caller to prepare a whole tip in advance. Instead, they’d have a pile of unsorted sequences, and during a tip they’d choose on the spot which sequence to call next. This is nice for flexibility reasons. I’ve already mentioned that a caller might want to choose the next sequence based on difficulty, but other considerations include avoiding calling similar sequences next to each other, or tossing out sequences with calls that have been taken off-list.

I have a pile of unsorted sequences, and I’ve called from it in the past. Lately I’ve been experimenting with preparing whole tips in advance. It’s natural when you’re only calling one tip at a dance, like I did for FCW. At the CCS dance I called, though, I prepared seven out of the eight tips I called in advance. I thought it went pretty well.

Some of the tips were themed around an idea. I had a C2 tip about the number three, a C1 tip I named “plenty of Plenties”, a C2 tip about the (non-)call Nothing, a C1 tip about “the most hated calls”, a C2 tip about The K, and a C1 tip about numbers that are more or less than usual. The call-themed tips might’ve been pushing it (the dancers were not amused after doing 40 The Ks), but I think the others were well-received. I also find it easier to write sequences about a specific idea, in the sense that it’s a creative limitation.

Planning a tip in advance also means I can make an arc. Pretty much everyone recognizes the importance of a short and easy opener, and the importance of ending at a sequence that people dance successfully. You can do more if you plan a tip, though. You can end a tip with a sequence that’s memorable or gets dancers talking. I like building up to a tricky or complicated application by rehearsing parts of it earlier in the tip. It’s not quite story-telling, but you can definitely craft a narrative.

Other nice things about working at the tip level: it let me manage the “idea density” of a tip easily, and it let me see which calls were used more or less at a glance. The main drawback is losing flexibility. I addressed this somewhat by writing two or three extra sequences to swap between, which felt like enough leeway.

Dances. A tip itself is usually part of a dance with several tips. Most dances I’ve been to are around two-and-a-half hours, and have seven to ten tips. Usually, a dance has a floor level, and most tips are called at that level, but I’ve also been to dances that alternate between two or more levels. Sometimes there’s a star tip, which is called higher than the floor level. I haven’t thought too much about the structure of dances, since I’ve only ever called three of them, but I might do so next time. There’s structures at a higher level than dances too, like weekends, conventions, or dance series. I might have to worry about these when I start teaching a square dance class, but probably not before then.

Kinds of sequences

I try to have some variety within my sequences, even in themed tips. Part of it is call-level variety, which means using the range of calls on a given list. Part of it is get-in variety, like balancing Heads and Sides, or doing something different every once in a while. Part of it is get-out variety, like balancing Right and Left Grand, Allemande Left, Dixie Grand, Promenades, and at-homes. But there’s also variety in the type of a sequence.

All sequences can be judged along several axes. How does it feel to dance? How surprising is it, and in what ways? How difficult is it, and what kind of difficulty does it have? I’ve found that dancers like having sequences that vary along each of these axes, even in a tip that’s meant to be flowy or puzzly. Some notes on these axes:

Dance feel. Consider the following sequence:

For most of these calls, the dancers all move at the same time. (The exceptions are the Heads Slide Thru at the beginning, and the last part of Swing and Mix, where the Ends wait for the Centers Trade.) The calls are all common for their level, and are called from familiar formations. The Turn and Deal starts with dancers turning in the same direction they were turning from the Peel Off, so there’s good flow. The Spin the Windmill ends with the Centers casting by the right, and with the Swing and Mix, they arm turn the ends by the left; this is good hand alternation.

These result in a sequence that, in my mind, is particularly flowy and feels good to dance. There’s also room for variety in the kinds of things that make a sequence flowy. Things that don’t feel flowy, in contrast, are calls where dancers have to think a bit before moving, calls with a lot of lateral movement, or calls where dancers stand still for several beats. The concepts As Couples or Tandem tend to break flow, more than concepts like Couples Twosome or Tandem Twosome. None of these things are bad on their own, but they contribute to making a sequence feel less flowy.

The kind of calls I’m not as sure about the “biggies”. These are the long eight-person calls, like Relay the Top, which is 16 beats, or Spin Chain and Exchange the Gears, which is 26 beats. Many of these calls are dancer favorites, and in theory, they’re supposed to be nice and flowy calls. In practice, I feel they can break the flow of a sequence, especially if they come after a rapid sequence of calls? I’d like to hear how other Challenge callers think about these big calls.

Surprise. A sequence can be surprising in several ways. Some examples:

• The same call meaning a different thing. From facing lines, Split Circulate 1-1/2, twice.
• A familiar call used in a neat way. From a normal trade by, Girls skip the last part, Tally Ho; Galaxy Circulate.
• A surprising get-out. From an I, with a left-hand wave inside a box facing out, Plenty; Slide Thru can end at home.
• Introducing asymmetry, and then resolving, especially when the calls that restore symmetry aren’t obvious.

What counts as surprising varies by dancer. I no longer find the first example surprising, but I was amused when I danced the second example two months ago.

My favorite kind of surprise is humor. I love sequences that are funny or amusing. Some more examples:

• I called an asymmetric, and the dancers ended up on the far 8 of a 1×16 formation, facing me. I then called the non-call Recenter, and a few dancers laughed as they walked a few steps forward.

• Andy Tockman called a C4 sequence with the excellent call Checkpoint Right and Left By By By By Touch By 1/4 By Single Right and Left By By By. Part of what made it funny was delivery; they said it without any parentheses and rattled off the bys in a rhythm. (Of course, they then said the call again, with pauses, to clarify.)

• Justin Legakis called a sequence that had something like “Swing the Fractions, but replace the first part with some call. Go and do that much, then wait. Replace the second part with some call. Replace the third part with some call. Replace the fourth part with some call,” followed by an extended silence. Eventually the dancers realized they had to do the fifth part of Swing the Fractions, and people laughed.

My favorite thing about sequences like these is that dancers remember them. At FCW, I had this conversation about the round dance El Mismo Sol:

I considered Della’s comment to be high praise, and I’m glad that dancers find my calling funny. People don’t square dance to hear a laugh-a-minute stand-up routine, but I think a little surprise goes a long way. I’m guilty of cramming too many surprises in a single tip, though; I’ll spread it out a bit in the future.

Difficulty. Part of the reason people do Challenge square dancing is to, well, be challenged. Difficulty can arise for different reasons:

• Requiring dancers to know the exact definitions of calls can be difficult, especially for calls that are often danced “by feel”. Consider Crossfire from a facing diamond, which ends in a T-boned box. The ends Cross Fold and stop, and they need to resist the urge of holding hands with the centers. The fact that it’s a fast, 6-beat call doesn’t help.

• Working in a disconnected formation can be difficult. That’s why Concentric is a harder concept than Cross Concentric, because the ends have to work with each other, even though they’re far apart.

• Distortions can be difficult, especially when they can’t be “taken out”, like with Magic. A Magic Pass and Roll Your Neighbor and Spread is tricky to dance correctly, especially since the Spread has to be a Magic Spread.

• Working with phantoms is a staple of C3A and above, and it can be quite difficult, especially for the “biggies”. From right-hand waves, consider Split Phantom Waves Explode the Top. The leading center has to work with three phantoms for the Star 1/4 and Unwrap. That’s a part of that call that I often dance by feel, and I’m liable to mess up if I’m not precise.

• Then there’s the part modifiers, like Initially or Piecewise, which require not only knowing the parts of a call, but cooperating with your square to reevaluate the formation when necessary. Combine them with concepts like Echo or Reverse Order and it can get messy.

These sources of difficulty can also be stacked and combined in interesting ways. These kinds of difficulty are, to me, part of the appeal of square dancing, both as a dancer and caller. The plurality of my sequences fall in this category.

I try to follow similar principles for writing a puzzly sequence as I would for writing a puzzlehunt puzzle:

• You’re setting up the dancers to win. Or the mirrored phrasing, which I find more compelling: you are ultimately playing to lose. This I first heard from David Wilson. A sequence that no one dances is, most likely, not a fun sequence. Neither is a sequence that you have to excessively cue the dancers through.

• Have exactly one big idea. I first heard something like this from Foggy Brume: no ahas is dull, one aha is fun, two ahas is a stretch. Square dance sequences are more forgiving though, in that you can have no ahas and still have a fun sequence. But it’d probably have to be fun in another way, like being flowy or surprising.

• Keep it short. This, I think, is where my sequence-writing philosophy is more extreme than what I’ve seen from others. My puzzly sequences average around 8 calls long, and even my sequences in general average 11 calls long. A usual average might be closer to 14 calls. This is again partly inspired from puzzlehunts, where often the goal is to produce an elegant puzzle. Something something nothing left to take away something something.

The distillation of these principles lies in the CX tips I’ve called, where a typical sequence looks like:

There’s a get-in and a get-out that are as short as I could make them, with the sequence body being a single, difficult call. When I called this in FCW, I called this sequence twice, as only one square got it the first time. I added (what I thought was) a gentle amount of cueing the second time around, and all the squares got it that time.

That’s one reason to keep difficult sequences short: you can retry without using up too much time. Writing this sequence was somewhat of a success, as the dancers won the second time around, without too much fuss. But I consider having to retry in the first place a small failure. If something similar to an Echo Initially appeared in an earlier sequence, then this might’ve been successful first try.

Reasons for calling

Why call in the first place? How am I any better than a recording, or, dare I say, another caller? Why write my own material, instead of using someone else’s sequences, which I’ve done in the past? Why do I spend, on average, five times as long preparing a sequence than I do calling it?

Briefly: I want to call material that I’d enjoy dancing to more than I would for other material.

This isn’t to say there aren’t callers I love. There’s plenty of callers I enjoy dancing to, enough that I use my time off to pay hundreds of dollars to come to their dances. But I can imagine material that I’d enjoy dancing to more. Sometimes it’s about quantity. I like funny material and themed tips, which do exist, and I want more of them in the world. More than that, though, I want material that’s difficult for different reasons.

I wanted the kind of difficulty that you could have in C3A or C3B, the kind that came from knowing the parts of calls, nesting, and modifying them. The thing is, this kind of difficulty isn’t as accessible in C1, because there isn’t enough flexibility with the calls and concepts available. If you were a C1 dancer interested in this kind of difficulty, your only option seemed to be learning C3A, which would take months. But you don’t need all of C3A for that—if you chose things with care, you can get a lot of mileage.

And this brings me back to CX. I mentioned that the CX list was C1, plus a few calls and concepts from higher levels:

• the calls Hop, Latch On, and Sidetrack,
• the concepts Stable, Echo, Sandwich, Reverse Order, Add;
• and the part modifiers: Initially, Finally, Oddly, Evenly, Secondly, Thirdly, Fourthly, Piecewise.

In a previous iteration, CX didn’t have Add, and instead had Split Phantom Columns and Twosome. But there’s similar ideas behind both versions of the list:

• The additions are easy to describe and learn. I wrote out definitions for all of these in an email, and each definition was at most twenty words. The calls are all sequences of other calls. (Sidetrack is Zig Zag; Counter Rotate; Roll.) And the concept names are mostly self-descriptive. (Oddly [concept] [call] means apply [concept] to the odd parts of [call], for example.)

• There’s as large a calling space as possible, while minimizing the number of added calls and concepts. In practice, this felt like the main restriction. I wanted concepts that had wide applicability, like Stable or Add, which interacted in non-compositional ways. Think about how Reverse Order can combine with Echo:

  • Echo [concept] [call] means (1) Do [concept] [call]; (2) do [call].
  • Reverse Order [call] means to do the parts of [call] in reverse order.
  • Thus, Reverse Order Echo [concept] [call] means (1) Do [call]; (2) do [concept] [call].

CX was fruitful enough that I’ve written forty sequences of CX material. I can imagine calling a whole dance of CX, maybe even a whole series. And while it’s filled a niche in the kinds of difficulty you could create in C1, I still think you could do better. For example, you probably don’t need Sandwich, when you can use Add instead. And you can add other kinds of concepts that aren’t as readily available in C1: grouping concepts like Twosome or Single, phantom concepts like Split Phantom or Multiple Formations Working, and matrix calls like Loop or Shove Off.

I don’t want CX to be normative or anything. CALLERLAB exists in the first place because having standard lists for square dancing is important. And it’s not like I only ever want to call CX. I want to call funny sequences and gimmicky tips. I think variety is great, and I could get better at writing flowy stuff. But, and this is selfish for me to say, I want more of the square dancing I like. There’s things I call that other callers don’t (yet). And maybe, when others dance to it, they’ll like it too. And maybe they’ll want to see more of it, and maybe callers will appeal to those interests. And then there’d be more of the square dance material I like in the world, and I’d be a little happier.