March 20, 2016 § Leave a comment
For quite a while I grappled with the issue of the “Subject”. Basically, the announcement of an essay about it once marked a temporary halting point in this proceedings. The outline once promised “The Self as an Issue of Temporality”…
Well, it took some efforts and time to leave some attached strings behind. I am quite happy indeed having met some triggers inside and outside, near and far, that collectively brought me to a field, or ground, that allows to re-start (cautiously, maybe). Not quite surprisingly, those were about the stance to the relationality sparkling among concepts, programming, and the programming of the ability to actualize concepts. That stance is also one about time, the time in language, the language about time, the unfolding of language in time, and the unfolding of time in language. Obviously, we remain affected by the issue of memory, and Bergson’s notion of durée is important for creating the crossing of different lines we have been developing in the past.
All the lines collected in the essays here started with the question about the limits or limitations of machines. Nikola Marinčić recently recommended to “leave the structuralist paradigm of computational models”, since “Every encoding is a form of stratification, where levels of abstraction on which we are able to stratify the encoding define the limits of our models.”
This statement certainly points to an important issue: The structuralist paradigm in computational models, let it be in computer science itself, or in any more application oriented field like computer-aided architecture\urbanism, is almost congruent with the model of the Universal Turing Machine (UTM). Following the UTM paradigm, nothing “new” can appear, nothing can “happen” in such an arrangement of encodings. A UTM-structured computer remains dead forever. Likewise, architecture following the structuralist approach must hope for its own failure in order to become interesting.
Despite the statement above by Marinčić points to an important issue, taking it literally it turns wrong: It excludes the possibility self-referential abstraction. Such not only does exist formally, in mathematics known from Category Theory. It is well possible to conceive of a material arrangement, intentionally created, or encoded, that can transcend its own being as mechanism. Such computing may be labeled “Non-Turing computation” as it extends its own axiomatic basis. Better perhaps we call it thus “Generative-Turing-Complexity-Computation”. Earlier we have been describing a possible route to that achievement, calling it Abstract Growth (see also the piece about design issues regarding the SOM).
When dealing with the limits of computing, one important, if not the most important part of structuralist approaches is their forgetfulness about time as durée, as becoming, and the event. (Just to avoid misunderstandings: in traditional computing the eradication of events within computations was or still is one of the main goals.)
Events are either open, both regarding their outcomes and their preceding underlying arrangements, or we would not deal with events. A closed event is not possible. The tic-tac-tic of the clock is not an event, engineering strives to annihilate any eventness not only in case of the clock. Purposes, functions, deterministic or linear algorithms, all of those exclude the occurrence of events.
Events may be observed on a field that provides a mediated situation within a temporary encirclement, though it is a surround without a middle. Events ultimately refer to the paradoxical logics of creation.
The key we would like to focus here is the observation. And this I want to take radically, at least for now.
Before we turn to the consequences, some examples may be helpful. Imagine observing a flock of animals you never saw before. Perhaps, if it helps to abolish your preconceptions, on another planet. In order to make any sense of them, you have to observe. In order to observe you need a measure. You need to calculate. To relate. To create a model. To test, and to expose yourself. Everything needs to be arranged by you. And finally you need to become like those, or even become them, in order to understand. You need to alienate yourself, split up. You could be bitten, indeed, if your conclusions were wrong.
All these parts and activities are inherent to any observation, and if you want to retain openness there is no other way to engage with the world than to observe. The alternative is simply to render the world into a trivial machine. Admittedly, we will meet this point again, there is no black-white either-or between those alternatives. It is more like a continuous struggling, or if more civilized, negotiating.
Observation is, temporally and logically, always a posteriori to the event. Inversely, without events there is no sense in observing. You cannot observe a clock. If you would, you implicitly expect its failure. As any talking about the event is even post-poned to the observation, any talking is post-poned to the event as well. Not just post-poned to the event a person is going to observe-describe, but also to the event of subjective language-formation.
Observation may be conceived as an instance of abstract measurement, the withdrawal of negentropy from the observed flow of entropy. We can now see that observation is a sibling of “meaning” in the Wittgensteinian way, since “usage” relates in the same way to the flow of entropy. Withdrawing negentropy from one context and transferring it to another one results in kind of re-formation of the receiving part. Just as Flusser mentioned, we get in-formed. (As a marginal remark we would like to add that in earlier times “informing” meant to “educate someone”.) Any activity, any action, any thought transfers negentropy, can be conceived as kind of an “observing usage”.
Whether we talk about observation or meaning does not make a big difference. From within the field of the event, both are aspects of the future. Yet, the concept of future with all its associated paradoxa related to “ordinal time” makes only sense if we switch to the world of causality. Nevertheless, we should not forget that information and causality are intrinsically linked, most significantly by the mode of observation, as we have seen earlier, summarized by this small chart.
Openness, both concerning the experience and its expectation, and such of course also its design in whatsoever context, implies a fundamental other-worldliness of the event, that always can be accessed only through observation, regardless where or how this openness should actualize or be actualized. Whether in a computer program, in a society as a matter-of-fact, or in a design exercise regarding a single piece of stone or a whole city.
Observation with its internalized modeling is mutually co-constitutively bound to this other-worldliness. The gap between them cannot never be bridged, yet the marveling void of the event gives roots to the construction of whole worlds. It is us that we become the other in this case too, meaning, approaching the event via observation we render ourselves into instances of openness. We also may suggest that in the fielded void of the event we can find, using a Deleuzean term, another de-territorialization of conceptual modeling. Or the other way round, we have to de-territorialize our thinking, we have to remove the mere possibility for a conceptual territory, if we are interested in the event.
Obviously, we cannot completely drop the encoding provided by standards, or the standardization provided by codes. Luckily we already know that standards and openness are not incommensurable in an idealistic way. They are linked by, or related in complexity. In language this interaction between codes, codifications, structure and openness, linking and linked by its autonomy and post-poned interpretation, is part of our everyday experience.
Now we (again) want to repose the question about the limits of machines. First of course we need to abolish the machine. So let’s refine the issue: What are the limits of the machinic? Or in more general terms, introducing a twist by inverting: Is it possible to implement the conditions for the possibility of the event?
Expressed in a contrasted manner that would render into:
What are the conditions for the possibility of encoding observable deterritorializing openness?
An immediately visible condition is the durée (in the “implemented” encoding): differential memory-time from which time and memory derive. The capability to refer to a difference in time, that is possible only through the capability to (voluntarily) select and apply an arbitrary encoding suitable for observation. It is not yet story-telling, instead we may call it story-drawing.
Interestingly, the openness of the event results in abandoning the directedness of the physical time arrow, similar to quantum physics. The act of observation, including all the choreostemic acts like encoding, modeling, etc. grabs potential determinations from the event, which in physical terms is before it. Seemingly a posteriori it even creates extractable negentropy and transfers it into the realm of encodings. Yet, overall there is no “before” or “after” in the relation of event and observation. Despite being distinct, they may intertwine and coil and braid around and through each other in parts of any proportions. Observations such inheriting eventness from what is being addressed. Unlike to the quantum in physics, however, there is no single smallest quantitative amount e.g. of energy in the event. The peculiar indeterminacy of the event may appear on any macroscopic level. (Also beware of invoking infamous “quantum computing” here)
Yet, let us multiply this active transition from some anything (flow of indeterminates) into any something (dispersed items), literally this decoherence, and we quickly reach a state of story-telling. Maybe story-telling is just the label for such braided, intentionally initiated decoherence of events into encoded items. Significant enough, any particular decoherence can not be repeated. Any of them is purely singular.
This way, activities like episteme or language are transcendentally conditioned by the event and its openness and their decoherence. This is the reason, why machine-based episteme must distribute itemized data into organized randomness first-hand. Without it, there would be no events possible. Obviously, the randomness cannot be white-noise, it needs to be organized. Yet, we also can see that there are no representational data any more. Representation by itself prevents effectively the possibility for events, ie. for episteme, even for the simplest.
Roughly spoken, the line of argument for the issue of the “subject” would be broadly the same. Maybe, subjects could be conceived best as self-maintaining story-telling stories, playful multiplication of decoherences of the same voids, each of the stories viz subjects perfectly distinguished from any other, forever, and even in parallel universes.