I agree here – the notion of ‘completeness’ and the Church-Turing thesis concerns universal representability. [I still think the variety of operations which make ‘completeness’ possible are hard to beat. but then there are pure analog parts which could extend the framework – linear parts, metastability – that could be of interest.]
Of course, actual Turing machines also transcend representation. For instance the Ackermann function has no other representative purpose – it is its’ own representation (i.e. the outputs grow so large that the only way they can be represented is in terms of Ackermann functions of lower orders). Even writing it down or describing it involves a Turing type framework.
For me the interesting thing about computation is not what can it represent, but what new kinds of objects and ideas come forth when a human is in the loop with or somehow forms a network/rhizome with its basic operations. The visual works of Vera Molnar, Frieder Nake and Manfred Mohr are instructive, as are once popular but fading concerns such as cellular automata and fractals. As are Myron Krueger’s early experiments in ‘artificial reality’, which do involve representation in some respect, but not in this more restricted and literal sense we have in modern VR/AR which is more in line with Kevin Kelly’s ‘mirrorworld.’
So I don’t the interest here is tied to any one medium – visual, auditory, etc. Computing is a medium on its own terms. It’s interesting to trace something like the rejection of pictorialism in photography – if there are those who would equally fight for computing on its own basis.
Here’s where I start to diverge – to see computing as a medium on its own basis – then the sense of mediation; ie. the connection (interface) between user/programmer and operations certainly belong to this medium and would have a great deal of influence. Conventional modular systems and desktop computing interfaces are then both very specific ways of bringing forth this medium and influencing what it can do.
The reason I brought up Leo Kupper is he seemed to have invented a completely different approach to the medium. (Is it even analog or digital?) It’s still very difficult to get documentation on this system – but it’s as if from another planet: https://www.facebook.com/asbl.artem/posts/225757787629438/ [sorry for Facebook link, difficult to find something more direct]
Circuit bending (the specifically digital explorations of Reed Ghazala, not the earlier analog effort by Michel Waisvisz that should be better known) is yet another approach. To say – we have logic structures at the circuit level but we will introduce feedback and make the even lower level aspects of the circuitry – ‘abstracted away’
Live coding is also very much in this spirit, and the Orca discussed here I think has helped it mature to a new level, because it takes away layers of representation.
There’s also a great recent article on Hyperallergic, in which the constraint of small program size brings out applications of computing that are in a sense most ‘its own’ – that is most removed from representation. Sebastian Bartlett’s piece was particularly affecting, it makes me think how I can with a few modular sequencers and logic gates perhaps attempt to evolve similar structures…
The article mentions hacking/hacker aesthetic, which has also an important role. We must become aware of the professionalization of programming practices which encourage the reusable, the generic, the correct (that is the representative) and ask – does our unthinking adoption of these practices, do our unquestioned ideas of ‘better code’ lock in certain biases or lead only to certain results (such as the endless-recycling loops of virtualization and simulation?) Look at most game development which is simply rearrangements of high-level modules. Whereas a lot mentioned in the Hyperallergic piece targets only the lowest of the low level and is certainly not ‘better code’.
Also, does our focus on the end result as music, as painting, as video etc. (interestingly, never ‘as experience’ – which was Krueger’s goal) also detract from the various ways computing can come into its own as an autonomous, non-representing medium which is always its message?
Lastly there’s sort of a manifesto for this perspective – Friedrich Kittler’s “There Is No Software” – well worth a read for any computer scientist, and kind of a hacker manifesto in some sense. I believe it contains a passage which floored me, about Alan Turing himself learning and preferring to think at the level of hardware (so he as a subject was himself transformed by the interaction of programming, which could thus no longer be thought in terms of representation). Could be another one of Kittler’s essays, perhaps “Protected Mode.” I think it could be well over 30 years old.
