qiudanz technique

qiudanz technique is a movement practice based on the computational transformation of movement sequences. its purpose is to create and share dances based on abstract computational machines. it is part of a commitment to preserve computer science beyond electronics and industrial civilization.

it is a technique developed and practiced by the txiemonks.

txiemonks

photo of a smiling person who seems to be dancing, with extended arms in front of them. they are pointing towards the left, where there's an overlay of some kind of code written with addition and subtraction signs, periods, and square brackets.

these are notes in progress. see the qiudanz devlog for updates.

qiudanz devlog

we have been developing qiudanz mode 1:

qiudanz mode 1

video: intro to qiudanz technique mode 1

movement sequence

we will be working with linear sequences of movements from a given and constrained vocabulary.

the first movement of a sequence is the "head", and the last one is the "tail".

these sequences can be of an arbitrarily large length, depending on the skill and practice of the participants.

modes

four potential modes so far, that correspond to the size of the vocabulary and of the number of transformations:

mode 1: two movements and transformations
mode 2: four movements and transformations
mode 3: eight movements and transformations
mode 4? sixteen movements and transformations

current notes

qiudanz mode 1

transformations

these transformations can be represented with the corresponding movements of the given mode.

mode 1

DEH: remove movement from the head
IN x: append movement x as a new tail

these transformations are enough to perform tag systems, cyclic or not.

mode 2

DEH
IN x
ROL: circular shift right: move tail to head
LER: circular shift left: move head to tail

these transformations should be enough to perform (more easily) the tape of a turing machine. máquinas de turing

máquinas de turing

mode 3

DEH
IN x
ROL
LER
INC: increment movement at the head
DEC: decrement movement at the head
INV: invert movement at the head
SWP: swap head and tail

more interesting possibilities!

movements

the size of the vocabulary of movements depends on the current mode.

these movements in the vocabulary have an order: there's a first one and a last one.

operations

to increment a movement implies converting it to the next one in the vocabulary. if the current movement was already the last one, then incrementing it converts it to the first one.

to decrement a movement implies converting it to the previous one in the vocabulary. if the current movement was already the first one, then decrementing it converts it to the last one.

to invert a movement one should divide the current vocabulary in its middle to get two parts of the same size. the inverted movement corresponds to the movement in the other half that is at the same distance of the middle than the current movement.

modes

pending.

the movements might be based on the work on choreutics.

choreutics

performance/play types

solo: perform a new movement sequence each time by applying a specific transformation to the previous one
solo card: the same as before, but taking the transformation from a card or other randomizer
conversation: one person performs a sequence, the other repeats it with an applied transformation, and this continues back and forth. the receiver could perform the original sequence (optionally synchronized with the transmitter) before the transformed one.
card conversation: the same as before, but taking the transformation from a card or other randomizer
tape and guide: one person performs and transforms the movement sequence continuously, as indicated by the transformation(s) given by the guide in the form of movements.
target: starting from a given movement sequence, apply succesive transformations to get to a target sequence. if playing in the tape and guide configuration, try to have the guide memorize the transformation sequence as its own movement sequence :)

a way of complexifying any of these would be to learn to apply more than one transformation when/before performing the new sequence.

another possibility would be to have beforehand a list of transformation to apply, with some or all of them being conditional on the movement currently in the head.

turing completeness

having some way of assigning the transformations to apply given the current movement in the head, would allow to perform arbitrary tag systems and/or turing machines, and m>1 tag systems are turing complete, e.g.:

Universality of Tag Systems with P = 2 (1964) Cocke and Minsky

however, cyclic tag systems are also turing complete, and they only require the sequential application of rules:

Cook, Matthew (2004). "Universality in Elementary Cellular Automata" (PDF). Complex Systems. 15: 1–40.