No More Learning

Decorum could once more assert the divisions of segmentation. But toward the mid-seventeenth century, the notion of deco- rum dissolved, to be recast in the form of die contract, which was the only way to protect the social order against the danger that people might change their "person" and turn out to be other than what they appeared. What re- mained oiimitatio in the old sense now referred to human emotions and to the impression created by the unusual despite its recognizability.
What we retrospectively describe as "baroque" is in many respects a combination of church directives and a sense of art aiming toward auton-
157
omy, toward form.
stasy, and heroism, which could be exploited equally well for religious and artistic purposes, paved the way for a rapprochement. Church-political measures that sought to influence the artist via legal and organizational constraints, supervision, and force led to artworks that, in retrospect, were nevertheless classified in art-historical terms as expressing an artistic style. Even in the more restricted domain of church painting, one finds a degree
158
of technical expertise likely to raise doubts about its religious inspiration. The state-political interventions of the twentietli century did not repeat these measures. Political attacks on modern art confronted an entirely dif- ferent situation. The autonomy of art had been historically established; it was now part of a history that lives through art, either by continuing the tradition or, more typically, by turning its back on it, by overthrowing tra- dition in search of new beginnings. In order to forestall the internal dy- namics of art, one can resort to political force and permit only politically correct productions that no longer impress the art system. Society has set- tled for autonomous functional systems, however. In the meantime, the art system has discovered an antidote to infringements by religion, poli- tics, or industrial mass production: namely, the distinction between art
and kitsch.
Secondary motives such as eroticism, asceticism, ec-
? 5 Self-Organization:
Coding and Programming
I
We speak of self-organization whenever an operatively closed system uses its own operations to build structures that it can either reuse and change later on, or else dismiss and forget. Computers depend on exter- nal programming, although computer-generated programs may be devel- oped eventually. By contrast, autopoietic systems produce their own struc- tures and are capable of specifying their operations via these structures (structural determination). This mode of operation does not exclude causal environmental influences. Some of Munch's paintings bear traces of water damage because they were left outdoors. While some people might consider this beautiful, no one would argue that the rain completed the painting. Nor would anyone try to prove the appropriateness of the rain's decisions with regard to the altered formal structure of the painting. Rather, the impression is that a painting was not and could not have been painted in this manner.
Self-organization owes its possibilities and its room for play to the dif- ferentiation of the system. Accordingly, art observes itself by means of the distinction between a reality "out there" and a fictional reality. The dou- bling of reality generates a medium of its own, in which the fixation of forms becomes not only possible but necessary, if the medium is to be re- produced. The opportunity and the need to do something go hand in hand. This conceptual model will guide the following analyses.
In functional systems, we call the system's basal structure--a structure that is produced and reproduced by the system's operations--a code. In
185
186 Self-Organization: Coding and Programming
contrast to the concept of code in linguistics, we think here of a binary
schematism that knows only two values and that excludes third values at
1
the level of coding. A code must fulfill the following requirements: (i) it
must correspond to the system's function, which is to say, it must be able to translate the viewpoint of the function into a guiding difference; and (2) it must be complete in the sense of Spencer Brown's definition, "Distinc-
2
tion is perfect continence," rather than distinguishing just anything. The
code must completely cover the functional domain for which the system is responsible. It must therefore (3) be selective with regard to the external world and (4) provide information within the system. (5) The code must be open to supplements (programs) that offer (and modify) criteria to de- termine which of the two code values is to be considered in any given case. (6) All of this is cast into the form of a preferential code, that is, into an asymmetrical form that requires a distinction between a positive and a neg- ative value. The positive value can be used within the system; at the least, it promises a condensed probability of acceptance. The negative value serves as a value of reflection; it determines what kinds of program are most likely to fulfill the promise of meaning implied in the positive code value.
Whether "tertium non datur" holds for the logical analysis of artworks
3
as units is debatable (but this is true for any unity of distinct objects). Ac-
cording to Kristeva, the work of art either does not exist at all, or it is a processing of distinctions--it is either a "zero" or a "double," but not a simple unit that can be negated in a single instance. This formulation may be premature, for one can certainly negate a double or treat it as a basis for excluding third possibilities. The question of how autonomy can be un- derstood in logical terms leads further. Whenever a system (or a work) claims autonomy, it must entail the possibility for negating autonomy; in addition, it must be able to negate this possibility. If the system claims more than autonomy, if it wants to observe and describe itself as au- tonomous, then it must take additional precautions to ensure that its code
4
is accepted rather than rejected. This is because society anticipates a vari-
ety of differently coded functional systems and therefore can operate, as society, only "poly-contexturally. " As we shall see, this condition affects the classical status of the idea of "beauty," which does not distinguish be- tween disjunctional and transjunctional operations. As a result, the differ- ence between "beautiful" (positive) and "ugly" (negative) is grounded in the idea or the value of beauty itself, which implies that the beautiful is simply beautiful.
Self-Organization: Coding and Programming 187
For the time being, we shall deal with straightforward binary coding. According to our understanding, a code is a structure among others--a structure that makes it possible to identify operations as belonging to a certain system but is not necessarily capable of representing, without para- dox, the unity of the system within the system. The question remains whether the art system is aware of a code that permits it to recognize what is or claims to be art and what is not art.
Codes are distinctions, forms that serve as observational devices. They are mobile structures that are applied differently from situation to situa- tion. To speak of a code is not to make a claim about essences. No matter which words we employ to describe the code (we shall return to this point), the code, by virtue of its binary structure and its closure, plays a crucial role in the differentiation of functional systems--just as the yes/no code is indispensable for the emergence of society. The primordial signif- icance of the code has been concealed by an obsession with "hierarchy"-- which advanced civilizations tend to regard as a prerequisite for order-- and by an acceptance of bivalence that has been reduced to the level of mere logic. Even today, the notion of the code still needs clarification.
The function of codes is to symbolize and, at the same time, to inter- rupt the basal circularity of self-implicative autopoietic systems. The clas- sical objection against tautologies, the petitioprincipii, is rendered obsolete by this insight. Within the code, the short-circuit of self-reference is sym- bolized and at the same time treated as a specialized phenomenon. Nega- tion requires a positive operation of "crossing" or "switching," a position that equals a negated negation. The code contains itself and nothing else. At the same time, the differentiation of two values interrupts circularity and creates asymmetries--in short, it generates systems. One needs addi- tional information in order to distinguish between positive and negative values. It is possible, in other words, to condition the system in such a way that it can decide which values are to be selected under what conditions. Such if/then conditionings (which are subject to exception or interpretive
5 constraints) lead to the emergence of self-organizing systems. In abstract
terms, the code is nothing but an invariant disposition for interruption. On the basis of this disposition, which is always given with language, the system can grow and increase its internal complexity in a historically irre- versible manner--initially through chance events and later on the basis of self-organization.
In the process of rendering itself asymmetrical (which exploits rather
188 Self-Organization: Coding and Programming 6
than neutralizes circularity) the system generates time. Time is necessary to cross the boundary between two values. The operation that accom- plishes this crossing takes time. Once the code is established, the implic- itly assumed temporal dimension unfolds into an explicit observational schema. The system needs a memory in order to know its initial condi- tions at any given time. It establishes a bipolar stability that causes the sys- tem to oscillate continuously between both of its values, and that keeps the system's future open by refusing to settle for one of these values. In ob- serving itself, such a system relies on an operatively actualized present that allows it to distinguish between past and future.
It is inconceivable that an art system could emerge without coding. Moreover, the code allows the specificity of a system--in this case, the art system--to be indicated by distinguishing its unique code from the codes of other systems. Codes respond to the problem of how to identify oper- ations as belonging to the system and therefore require certain qualities-- this distinguishes codes from other types of distinctions. Above all, codes must be formulated abstractly enough to inform every operation in a given system. The code's reusability must function as the equivalent of the indication of the system's unity, but without obstructing the functioning of the code's mobile structure. The code exists only when it is used to con- strain the recursive anticipation and recapitulation of operations in the same system. Compared to other functional systems, the specificity of the art system resides less in the names of the code values than in that the task of rendering these values asymmetrical (via conditioning and via the gen- eration and use of time) is primarily the responsibility of the artwork it- self, and that intermediate levels--such as rules or concepts of style--are possible, but to a large extent dispensable.
Despite the difficulties in finding convincing names for the code values of art (for example, values analogous to the true/false distinction in sci- ence), distinguishing between problems of coding and problems of refer-
7
ence and their corresponding distinctions is of utmost importance.
lems of reference occur whenever one distinguishes between self-reference and hetero-reference or, in our case, between art and nonart. The unity (form) of the distinction self-reference / hetero-reference supplies the sys- tem with a satisfactory conception of the world, thereby concealing the dif- ference between system and environment that is initially generated in the operation. Problems of coding, by contrast, concern the difference between positive and negative values, which the system uses to indicate which op-
Prob-
Self-Organization: Coding and Programming 189
erations belong to the system. Problems of coding divide the system's self-
reference along the lines of what is acceptable and unacceptable, that is,
they always refer to the system itself. So far as the environment is con-
cerned, acceptance is not an option. The environment is the way it is, or,
put differently, the system has no freedom in the environment. Distinc-
tions of reference and distinctions of coding--we are always dealing with
distinctions! --are positioned orthogonally in relation to one another. This
is why reference to the environment cannot function as the negative value
8
ofthecode. Distinguishingaworkofartfromsomethingelsecanonlyde-
marcate an observational space and signal that within this space observa- tional relations of a special kind hold, even though it may appear as if the environment is admitted into the system--in the form of a bathtub, for ex- ample, or as a noise audible when the music stops playing, or as an almost normal newspaper advertisement.
The art system must be coded. It must be able to presuppose a code of its own that cannot be surpassed within the system; otherwise it would be un- able to differentiate artworks as a special observational domain. This is true even if the art system's self-description were to orient itself, in a more tradi- tional fashion, toward "principles. " It would still have to decide whether a given work is or is not in agreement with these principles. If everything is acceptable, then it becomes impossible to distinguish art from nonart. Without this distinction, there is no way to dismantle the distinction itself. In order to create possibilities for observation, one must begin with a dis- tinction; if these possibilities are to be specific and distinguishable, then one needs to begin with a specific difference.
The differentiation of the art system manifests itself in the indepen- dence and distinctness of its coding. This is evident in the relationship be- tween aesthetics and morality, in contrast to the moral obligation of art
9
that still held for Gottsched. Sexual morality provides a poor testing
ground in this regard, because morality, in this domain, is in a state of flux (despite criticism from the French side). The extensive theodicy debates in the wake of Leibniz and Voltaire's Candide were more likely to show how badly moral justifications were in need of reform (a problem to which Adam Smith, Kant, and Bentham responded accordingly) rather than in- dicate a conflict between the codes of art and morality. But the problem is
10
also considered more as a matter of principle.
necessarily have to agree with the morally good, and it cannot draw sup- port from such agreement if it doesn't convince as art. Yet there is no such
The "beautiful" does not
190 Self-Organization: Coding and Programming
thing as a crosswise identification of codes, as if the beautiful would have
to prove itself above all in the realm of moral perversities (such as incest).
The distinctions are positioned in an orthogonal relation to each other;
they are indifferent to one another. But since one sought to distance one-
self from a tradition that thought about this problem differently, the for-
mulations expressing this trend tend to be uncertain and ambiguous.
Friedrich Schlegel, for example, writes, "Isn't it true that a certain [! N. L. ]
aesthetic malice is an essential component of a balanced education? " And
he later rejects "the aesthetician's fashionable contempt--a contempt that
distinguishes nothing--of everything that is, or claims to be, morally
11
good. "
The problem is how to deal with the plural coding (or poly-con-
texturality) of modern society while still insisting on the unity of the (hu-
man) subject and on a bivalent logic. A possible solution is to recognize
and communicate codings as paradoxes, in aesthetics via irony, in moral-
12
ity, directly.
Demarcations of this kind are of little help when it comes to determin-
ing die values of codes. In traditional aesthetics, the code values of art
13
were labeled beautiful and ugly. Nevertheless, aesthetics tolerated the
artistic depiction of ugliness. As early as the Renaissance, storms, fires, and so forth were much admired topics of painting (later on, they would be called "sublime"), and in order to paint such things, one drew on the very principles one followed when drawing beautiful objects (without altering the technique of perspective, for instance). In Herder's words, ugliness
14
played the role of an "ancillary idea. "
in a double sense (and, in this regard, paradoxically): it was opposed to ugliness and it implied a general judgment about the relationship between the beautiful and the ugly, or, put differently, beauty was applied both at the figural level and at the level of the artworks unity.
This is why one was not in a position to distinguish between coding and programming. The level at which art represents objects was not dis- tinguished sharply enough from the level of coding, even though the rep- resentation of ugliness, evil, and deformity was justified as a contrast, that
15
is, in view of the other side of the difference beautiful/ugly.
that, the principle of imitation suggested that art ought to represent both
16
kinds of objects.
so on, one was thinking of the nexus between the parts of a whole rather
17
than of a manner of operating.
no more than a transgression of the medium's possibilities,
The concept of beauty was applied
And when discussing "appropriateness," "fitness," and
Lessing already considered ugliness to be
18
reserving the
Apart from
Self-Organization: Coding and Programming 191
concept of beauty for the overall judgment of an artwork. Beauty, in other words, was understood to be a viewpoint of judgment rather than an op- tion that emerges along with the process of artistic production. This is how the contrastive formula beautiful/ugly is still applied in romanticism, despite the trend toward explaining ugliness, the negative side of this dis- tinction, in terms of a "crude" taste or a corrupted moral sense, or in sim-
19
ilarly disapproving ways.
It has become increasingly difficult to insist on the labels beautiful/ugly
for the positive and negative code values of art over the persistent protest
20
of the system.
only to artworks but to other objects as well--for example, to people. order to maintain this parallelism, the beautiful/ugly distinction must be restricted to the figurative level and therefore fails to account for opera- tions of observation (such as producing and observing an artwork), which, as operations, are neither beautiful nor ugly. The problem seems to be that this distinction refers the criteria for judging artworks to descriptively com- prehensiblefeatures of individual works, or, conversely, that it infers from such features criteria that can be generalized. Under such conditions, it is impossible to separate the levels of coding and programming in a manner characteristic of the functional systems of modern society and their "posi- tive" programs.
One therefore wonders whether the values "beautiful" and "ugly" were ever meant to serve as code values to begin with, in the sense of indicating a relationship of exchange mediated by negation. Be this as it may, the tra- dition was unaware of the distinction between function and coding as a way of distinguishing between different forms of specifying a given system. In the idea of beauty, both aspects converged, and even in the early twen- tieth century one still tended to define distinct arrangements of social life in terms of a priori values. One sought to describe social differentiation by differentiating between such values rather than exactly specifying a coun- tervail (or an external side of the form). Besides, it remained unclear whether the notion of beauty referred to individual parts (figures) or to the
22
beauty of the work as a whole.
of perfection pertaining either to the work itself or to its representations; beauty singled out the work, it served as a mark of distinction. Works of art simply are beautiful, or else they are not works of art. There was no need to distinguish art from failed attempts at art or from nonart. "Ugli- ness" could be integrated in the form of grimaces, dissonance, and so on
The reason might be that these labels are applicable not
The tradition thought of beauty as a form
21
In
192. Self-Organization: Coding and Programming
without distracting from the works beauty--in accordance with a general cosmology, which suggested that a world consisting of perfect and less per- fect beings--of angels and stones, men Wwomen--is more perfect than a world containing only the most advanced forms. While the principle of imitation reigned, it was easy to find a balance so long as art was allowed to represent both beautiful and ugly objects in accordance with the work's
23
internal ornamental structure.
Such a frame of mind suggested a notion
of art as idealization--of both beautiful and ugly objects--whereby
ugliness was tolerated as a means of emphasizing beauty by contrast; ac-
cordingly, one spoke of "the fine arts. " Along these lines, German Idealism
turned the notion of beauty into an Idea or an "Ideal" in which all op-
a notion that was still accepted without question in
Perhaps the notion of an ultimate value (and its identification with the code's positive value) was only a precarious transitional solution--pre- sumably modeled on the religious idea of God. After the search for "ob- jective criteria" of beauty had failed, the objectivity of beauty was con-
26
ceivable only as a tautology, as a circle --and the decision about which
forms convince as beautiful was left to history. In this way, one could hold on to a notion of unity, even though concrete forms rest on distinctions. At the same time, one could avoid conceiving the unity of the difference between positive and negative values as paradox. Hegel was perhaps the last to cast this idea into the form of a philosophical system. Today, any logical analysis would insist on separating positive and ultimate values by a difference in level. In Gotthard Giinther s terminology (which we intro- duced earlier), this means that disjunctive and transjunctive operations and their corresponding values must be kept separate. In this regard, the idea of beauty appears "confused" in retrospect, which suggests that its function as the ultimate value of the system was to conceal a paradox.
Defining the problem away in this manner is unacceptable. Instead, we must give up the notion that the art system's operations are structured tele- ologically, that artistic production strives toward an ultimate end, and that beauty is a criterion for permitting judgments about how a work ought to be made and whether it can please or not. This applies to all cases of bi- nary coding. In die true/untrue code, for example, the value of truth does not simultaneously provide a criterion for truth, as the venerable dictum
verum est iudex sui etfalsi [truth is its own judge and the judge of false- hood] suggested. Rather, the positive/negative structure of code values
posites converge, romanticism.
24 25
Self-Organization: Coding and Programming 193
must be distinguished from the criteria (or programs) that direct the proper choice of one or the other value. The positive value, in other words, does not promote itself; it is nothing but the inner side of a form that pre- supposes another side and cannot be indicated without this other side. In its traditional sense, the idea of beauty blocked the distinction between coding and programming, and introducing that distinction explodes this idea. If one wants to distinguish between coding and programming, then one must refrain from specifying the content of beauty (even if one thinks of it as the unattainable normative goal of an infinite striving).
Beauty, then, is neither a quality of an object (just as truth is not a qual-
27
ity of propositions) nor an "intrinsic persuader. "
The abstract bivalence
through which an observer observes artistic operations requires a third--
what Derrida would call a supplement--which simultaneously respects
and disregards the fact that the system operates under the logic of the ex-
cluded third. There is no way of supplementing the code by adding a third
value--for example, along the lines of "beautiful-ugly-tasteful. " Among
the criteria that regulate the evaluation of successful/unsuccessful opera-
tions, there can be a plurality of further considerations, which, however,
cannot represent the unity of the system as a form (a two-sided form, as
2
usual) within the system. When Derrida speaks of the supplement, * he is
thinking of the official status of such additions. Within the relevance hier- archy of self-organization, they are considered marginal. They fail to cover all the cases, and they do not apply to every operation or to the unity of the whole. This ranking can be "deconstructed" as a mere dictate of sys- tems logic, because, logically, the operation of the code requires supple- ments of this sort. In the terminology of Michel Serres, they are the "par-
29
asites" of the system --the included excluded thirds.
Starting out from the operations of the system, we can at least account for how a code comes into being and how it is used, how it works. Every operation--whether the artist's or the beholder's--must decide whether a given form does or does not fit, whether it can be integrated into the emerging work (or into the work one is about to inspect) in ways that se- cure connectivity. Every observation places the detail it indicates into the recursive network of further distinctions and, from this viewpoint, makes a judgment about the success or failure of this detail, thus distinguishing between solutions that convince and are immediately intelligible, on the one hand, and moves that are questionable, incomplete, and in need of correction, on the other. This is how a binary code works--which might
194 Self-Organization: Coding and Programming
include situations that require withholding one's judgment "for the time being" (the same is true in matters of truth). But without coding, there can be no decision; anything goes.
One must be cautious not to construe a fitting operation as a way of fa- cilitating the next move--as if we were dealing with a mathematical task or a technical construction. Adding further distinctions to the artworks form combination might facilitate connecting operations, but it can also make them more difficult. It might become questionable whether one will be able to continue, or whether one will ever manage to produce a work of art that is closed and rounded off. Often enough, the thrill resides in precisely this risk, in the lack of foresight, in the difficulty of the self-im- posed task. The only requirement is to stay clear of two limits: the neces- sary and the impossible. The artwork must remain within the modality of contingency and must draw its power to convince from its ability to pre- vail in the face of other self-generated possibilities. Suppose one needs a green color in order to balance two incompatible shades of red. But how about gray, which, if one were to use it instead, would look like green?
The fit or lack of fit, the success or failure of additions, does not affect the boundaries of the art system. Failed works of art are still works of art, if unsuccessful ones. This is why it makes sense to take on difficult pro- jects, to incorporate things that do not fit, and to experiment with possi-
30
bilities of failure. As the structuralists have taught us,
source for recognizing order. Likewise, the negative value of the code serves as a means of self-control, as a value of reflection. This is true even when art questions precisely this distinction--doing just that. If one wants to leave the art system, then one must look for orientation in a dif- ferent code or in no code at all.
O>O>O>O IIII IIII
23n Opi > Op > Op > Op
IIII IIII I>I>I>I
Operational sequences always transport both positive and negative con- notations without there being a way to recognize this fact except in the re- cursivity of operating--it cannot be identified as a telos or a rule. If this
trash is a prime
Self-Organization: Coding and Programming 195
proposition holds, it suggests that the code expresses an aggregate of these accompanying valuations. The accompanying diagram explains what we have in mind. The recursive sequence of observations condenses and asserts the system's code by presupposing it in each of its operations as a condition of both the operation's selection and the system's recursive recapitulations and anticipations. In this way, the code is enforced despite the variety of formal decisions. This happens no matter how aesthetics, in its capacity as a reflection theory of the art system, defines this code. There is still no con- vincing alternative to the labels beautiful and ugly. But this semantics is not to be construed as if art were concerned with "beautiful figures," "beauti- ful sounds," or other beautiful forms. If one wants to hold onto the se- mantics of the beautiful, then it might best be understood as a summariz- ing judgment about what fits /does not fit under added conditions of high complexity, that is to say, in the face of self-generated difficulties.
This does not yet explain how a code, as a component of the art sys- tem's self-organization, participates in the system's operations. Individual operations must be recognized as contingent; they must be "motivated" by the work.
This suffices as a condition of their intelligibility. Put differ- ently, neither the artist nor the beholder needs the additional specification "coded" in order to observe. Likewise, in research there is no need to men- tion that truth or falsity is at stake, apart from the theories and methods with which one is working. Invoking the code becomes necessary, how- ever, when the question arises of how art or science distinguishes itself from other functional systems in society. The specificity of coding repre- sents, at a level of third-order observation, the distinction between system and environment. This may be of practical significance if one wants to control the system's recursivity. A piece of raw nature or a slice of un- processed society might find its way into the artwork--a natural stone in a sculpture or a newspaper ad in a collage. But whatever is integrated in this manner must find its place. Its origin per se does not legitimize its participation in art. Incorporating items of this sort requires no reference to the code--it suffices to be aware what fits in a given instance and what still remains to be done in order to accomplish the integration. That, to begin with, this makes sense and is permitted, and that it can be observed as a distinct process presupposes a higher level of reflection that inevitably reflects upon the code. "Transjunctional" boundaries of acceptance or, more generally, the limits of possible aesthetic forms provoke a definition of art on the basis of its code. Accordingly, the meaning of coding is
196 Self-Organization: Coding and Programming
bound to become more and more abstract, if one can demonstrate that a work's form combinations are still within the realm of the possible.
II
The miracle of recognizability precedes all programs of art. It is an ef- fect of forms that have been distinguished. A figure can be recognized if it is perceived first in nature and subsequently in an artificially created con- text. A buffalo remains a buffalo when projected against the wall of a cave. Even material differences can be overcome in this way. A human head re- mains a head--whether it is cast in clay or in stone, whether it is drawn on a vase or on a wall. We can repeat and recognize a melody whether it is sung, whistled, or played on an instrument. Art consolidates identities be- yond what nature has to offer, and it does so with a certain indifference to situations, contexts, and materials. Art accomplishes both a condensation and a confirmation of form, thereby ascertaining the hidden order of the world. To borrow a phrase from the ancient Greeks, art allows a glimpse into the essence of things.
This must have been amazing at first. For millennia, the evolution of art has profited from art's capacity to expand and refine its formal reper- toire to accommodate recognizability and to detach its forms from natural models, at least to a certain extent. This could be accomplished without differentiating between coding and programming. Whatever "beauty" might have meant, it remained a matter of form, and form was commit- ted to striving for exceptional recognizability. The essential depths thus opened in the world could be considerably intensified; this was, perhaps, above all the accomplishment of the Greeks. But early forms of writing might also be understood in terms of this desire to fix certain contents and to secure their recognizability, especially when presented together with images, so that writing and image illustrate one another and facilitate the
31
recognition of a content that might also be narrated.
ily at stake in such representations was the problem of securing the world, and, in view of this purpose, a differentiation of art would have presented an obstacle rather than an advantage. So long as the certainty of the world depended on the recognizability of its forms, its hidden invariance and essences, aligning art with religion, with political power, and eventually
32
with aristocratic genealogies was a commendable course of action.
In the European tradition, a common basic idea was the notion of gen-
What was primar-
Self-Organization: Coding and Programming 197
eralization, which entailed the possibility for drawing, at the level of the universal, distinctions that constituted relationships of exclusion among the objects distinguished. Plato called what was so distinguished genos and the art of distinguishing it dihairesisP In the Greek word dihairesis (de- rived from haireo), it is impossible to distinguish among the activities of reaching out toward an object, of dividing, and of distinguishing the ob- ject. The g<<z<w-technique, one might say, constitutes a procedure for gain- ing access to the world that articulates and divides the world by means of distinctions. The basic rule is to avoid paradox. Although the genos com- prehends a multiplicity within the form of a genre, it is imperative that genres not be confused with one another. The to kata gene diaireisthai de- mands that one and the same genre cannot be another and that another genre cannot be the same as the first one. This is a requirement of cogni-
34
tion (episteme), which Plato calls dialectic.
ception of the ideas that allow for gathering many disparate things into one (despite their diversity). This technique, which Plato illustrates with
35
reference to grammar and the alphabet,
torical use of paradox; in other words, it is distinguished against the prob- lem of paradox.
The desire to exclude paradox constitutes the countertheory to rhetoric.
It combines the internal logic of writing, language, and technique with
the assumption that, by penetrating to the Ideas, one can arrive at an un-
derstanding of how the world is divided, and how one needs to place one's
distinctions accordingly. Technique is still understood to be an art of dis-
tinguishing bound by nature. Aristotle later supplies the concept of cate-
gory (= an accusation to which the world must respond) to indicate the
primary divisions of being. In his Poetics, Aristotle assigns to poetry the
task of representing the Possible (dynaton) as the Universal--that which by
necessity arrives at its destination unless it encounters an obstacle. To this
task corresponds the assumption that the recognition of cognitive insights
into the essence of things (insights that are difficult to come by) creates
pleasure, which justifies imitatio as the goal of art. The rhetorical notion
of amplification builds on this g^wof-technique. Rhetoric values amplifi-
cation positively, because it tests generalizations and retains successful
ones as "commonplaces. " This procedure was still common in the Renais-
36
sance.
the seventeenth century devalorized this tradition.
did not affect the general division of the world according to species and
The more stringent demands made on rationality and proof in
It presupposes a clear con-
is opposed to the Sophists' rhe-
37
At first, however, that
198 Self-Organization: Coding and Programming
genres, which prevailed until Kant explicitly raised the issue of the possi- bility for a future metaphysics, finding it difficult to appreciate the dis-
38
tinction between species and genres.
As a result of this genos technique, the concept of imitation could be
broadened significandy, if it not bent entirely out of proportion. Sir Philip Sidney (1595), for example, defines imitation as follows: "Borrow nothing of what is, has been, or shall be, but range (only reined with learned dis-
39
cretion) into the divine consideration of what may be or should be. " The demands of a learned (classical) education continued to constrain im- itation, while indicating the point where the concept is put at risk by fur- ther developments. "Imitation" seems to serve as no more than a cover for a differentiation of art that has already advanced considerably.
Whence the impulse to change this situation? Presumably, the motive
was an external incident: the loss and rediscovery of the artistic skill of
antiquity, which focused attention on how antique art was made. More-
over, print provided an opportunity to proliferate technical literature,
thus liberating artistic know-how from the oral doctrine taught in the
40
workshops.
primary--at first in the coordination of production and cognition (Ba- con, Locke, Vico). This move, however, only expresses what is implied in the technique of dihairetics and is later made explicit in Kantian theoret- ical technique, namely, the desire to pursue into the regions of meta- physics the question of how reality can be reworked by the subject. The transition from "what" questions to "how" questions always indicates a shift from first- to second-order observation, and second-order observa- tion now required programs of its own.
If this assessment is correct, then it should come as no surprise that the artistic programs of the late Middle Ages and early modernity appear in the form of recipes and rules. At stake in these programs is above all the renaissance of antiquity, the recovery of artistic skills along the lines of re- discovered thematic models. With the discovery of perspective in the late Middle Ages, however, second-order observation and the search for its rules went beyond that impulse to rediscovery. Learning the rules sufficed to master this new technique, and if following such rules led to deviations, the sheer demonstration of artistic skill might still count as art. The prime focus continued to be on recognition, albeit within an increasingly ex- panding, universal realm of meaning independent of thematic models. One studied the rules in anticipation, as it were, of everything that might
Questions of "how" gained prominence, indeed, became
Self-Organization: Coding and Programming 199
be considered art. In so doing, art severed its ties to religion, but in a way that included rather than excluded religious art.
Rules formulate a preference for doing things right. They are no longer abstractions of genos, but they are abstractions nonetheless. Rules antici- pate multiple applications in a variety of cases; indeed, this constitutes the regulative meaning of the rule without infringing on its identity. As be- fore, one avoided Plato's tautonlheteron paradox. To the extent that rules express certain preferences, they do not tolerate a distinction between cod- ing and programming. Following the rules was believed to be the condi- tion for a work's beauty.
A separation of coding and programming (and a reorganization of art in terms of self-organization) did not occur until novelty became an indis- pensable requirement of artworks and copying was no longer permitted.
and because it is not what was before. As Aristotle was well aware (Peri hermeneias IX), this explodes the logical law of the excluded third. Every- thing excluded must be condensed into a "third value," the value of unde- cidability. But how can this be, if one must eventually accept that the world itself becomes another world, a new world, from one moment to the next?
A new sense of novelty emerged from a covert revolution in the concept of time and an outspoken polemic against the Aristotelian heritage of the scholastic tradition. It affected the definition of time in terms of the ae- ternitasltempus distinction and the protection of beings within the pres- ence of the eternal. It concerned the presence of both the origin and the end, the actuality of the ground of being in every moment of its move- ment. Once one gave up these notions (which happened one by one in various thematic domains), a space was cleared for novelty, for a disrup- tion that requires meaning and selection. The only condition is that nov- elty must please, and practices of observation and description soon fo- cused on conditions of this sort.
Print might have been another motive for this "transformation of val- ues," especially cheap prints produced for the sake of entertainment and
42
Upon first sight, novelty is ontological nonsense. Something is, although 41
Novelty became a marketing strategy, since one could assume
polemics.
that no one who knew their content would buy such products.
The criterion of novelty settled an old sixteenth-century controversy. The
43
debate about demarcating poetry from science (or historiography)
had
raised the question of how poetry can expect to please if it represents false-
44
hoods and fictions. Apparently, only fools and children could take plea-
200 Self-Organization: Coding and Programming
sure in such things, and their use was therefore limited to educational con- texts. Temporalization, the shift of emphasis from deviation (from truth) toward novelty, created a respectable audience for poetry. As early as the six- teenth century, indications abounded that novelty was considered a neces-
45
sary condition for the surprising--and pleasing--effect of artworks.
first, this trend responded only to a specific problem in art and traditional poetics--the problem of how to account for art's interest in extraordinary objects and events {meraviglia in the widest sense). Because of the experi- ence of religious civil wars in the seventeenth century, innovation still car- ried negative connotations, especially in the domains of religion and polit- ical power and in the classical realms of natural and civil law (but not in the law of the "police," which was just beginning to emerge and claim its own territory). This is why tolerating--or even demanding--novelty could serve as a means of distinguishing functional realms.
Whereas antiquity valued the striking effect of certain objects as a con-
46
dition of recollection and a source of information,
the concept of novelty. One discovered the unique charm of novelty, de- spite its triumph over what was previously considered beautiful and often for this very reason. Sixteenth-century mannerism illustrates how this ten- dency became deliberate. (We shall address the question of "style" later. ) As a distinctive feature of art, novelty was introduced as a condition of pleasure--and it goes without saying that neither religion nor politics nor law must "please. "
The focus on "pleasure" or "pleasurable consumption" suggests that the relationship between producer and recipient, or between art and its audi- ence, was foregrounded in ways unknown in antiquity, in the Middle Ages, or even in the early Renaissance. Generally speaking, this trend strongly indicates that the transition to functional differentiation was un- der way, emphasizing a function-specific complementarity of roles every- where (buyer/seller, government/subjects, educator/pupil, lover/beloved). The distinguishing criterion of art, especially in its demarcation from sci- ence, was now sought in the manner in which art takes over the hearts of its audience. At the same time, the notion of "pleasure" appeals to indi-
47
viduals; it takes an individual to decide what is pleasing,
first not every individual was included--not every maid or peasant--but only those capable of judgment, individuals endowed with taste. In ret- rospect, one recognizes the transitional nature of this formula; it was a compromise that could convince only for a short time during the seven-
one now temporalized
although at
At
Self-Organization: Coding and Programming 201
teenth and eighteenth centuries. Only the individual endowed with taste could be stimulated by novelty, for only such an individual could decide what was new, even though one needed criteria to avoid falling prey to everything new.
The demand for novelty implies a retreat of time from all occupied
48
places.
placement, no need to prove superiority. To the extent that the principle of novelty takes hold, history and age no longer legitimate occupying places in a world whose sum remains constant. Novelty pleases because it doesn't need to be regarded as the outcome of some territorial dispute. Rather, novelty seeks to do justice to time itself by surpassing necessity through innovation. Novelty irritates in ways that resemble the delight one takes in paradox, but without proposing mere deviation from the or- dinary as a criterion for acceptance or rejection. Just like individuality, novelty challenges the bifurcation of the aristocratic world, of political ter- ritories and patron/client relationships, whose origin and age indicate that they have lost their meaning. Long before democracy will prescribe ever new elections, and long before individual destinies are made by careers rather than by social origin (as criterion), a sociostructurally harmless phe- nomenon such as art could place its bet on perpetual novelty. But how did art accomplish this? How can it bear the demand for novelty? Given the delight one takes in sheer irritation and provocation, how can one arrive at criteria for rejecting some innovations as failures?
On the one hand, one can form an art-specific preference for novelty (against copies). On the other hand, it is not feasible to code the entire art system along the lines of dated/new, thus devalorizing the entire stock of existing artworks--which one collects with great zeal. Nor is novelty suited to serve as a programmatic formula, since it provides no way of recogniz- ing which of the new works qualify as art and which do not. The difference between coding and programming solves this problem. The code remains stable, whereas what fulfills the program function of assigning the correct code value can be left to change, to the spirit of the age, and to the demand for novelty. The novelty postulate functions as a hinge that joins and sepa- rates coding and programming. Whatever else it may be, novelty is devia- tion. The novelty requirement destabilizes both the notion of deviation and the concept of rules. Valuing a work simply on the grounds that it fol- lows the rules no longer suffices; to the extent that one recognizes the work
49 was created by following the rules, it cannot be new and it cannot please.
There is no need for power struggles, no competition in this dis-
202 Self-Organization: Coding and Programming
The code requires abstraction in order to express a preference for an art that is valued positively, and this is why one cannot derive rules from the code for the proper production and evaluation of artworks. And since new works are constandy being produced and exposed to evaluation, the possi- bility of a program of art that would not be cast in the form of rules be- comes questionable. In a sense, the doctrine of taste was the last attempt to answer this question in the affirmative.
Ill
The task of distinguishing between coding and programming requires distinguishing distinctions--not only objects, rules, or points of view. Ab- stract coding, which identifies a given operation as belonging to the sys- tem, already distinguishes between a positive and a negative value or, in traditional terminology, between beauty and ugliness. But this only se- cures the general contingency of the system's operations. In addition, one must be able to distinguish, at the level of programming, between correct and incorrect assignments of values to the code. In other words, it is pos- sible to apply the system's programs incorrectly, even though mistakes of this sort do not automatically attract the negative code value in the sense that they would be perceived as ugly. In the doctrine of good taste, these two evaluative levels are not clearly distinguished. A venerable truism states that even ugly objects can be represented artistically (although the
50
theory of art has a difficult time accepting this proposition ). Moreover,
the doctrine of good taste does not base its evidence on criteria but on the fact that there are clear-cut cases of bad taste. The question is: Can an art- work fail without therefore being ugly? And if so, how?
One might speak of failure when an observer loses control over a work's play of forms, when he can no longer understand how a particular formal choice relates to others on the basis of what this choice demands of the work as a whole. But this can be demonstrated only with reference to a concrete case, not by applying principles or rules.
We might answer this question by considering that every artwork is its own program, and that it demonstrates success and novelty if it manages to show just that. The program saturates, as it were, the individual work, tolerating no further productions of the same kind. At a conceptual level, this excludes the case on which Arthur Danto focuses his aesthetic theory, namely, two objects that look identical and are aesthetically indistinguish-
appear as constraints. "
Because it is difficult, on this basis, to account for
Self-Organization: Coding and Programming 203
able but are "transfigured" via interpretation into two different works of
51
art.
interpreted differently. ) Serial painting, which experiments with different versions of the same image, might also be acceptable. But this is a varia- tion on the basic idea of the self-programming artwork--a variation that, compared to a work confined to a single location, permits the display of greater complexity.
The Kantian formulation conceives the artworks self-programming as the freedom of the observer to let his cognitive faculties play without be- ing guided by concepts. The point in speaking of an "end in itself" or of a purpose without purpose--for Kant, at any rate--is to distinguish art
52
from a conceptually fixed cognition.
ters what we call self-programming without explicitly naming the phe- nomenon. Kant starts out from the cognitive faculties and subsequently directs his critical efforts against every position previously occupied by metaphysics. In this context, art has barely a chance to articulate itself-- except by stretching traditional terminology in ways that, even prior to ro- manticism, were not considered very helpful. What remains noteworthy is that the concept of freedom problematizes the observer, while blocking further pursuit of what the observer's function and role might be with re- gard to a self-programming artwork.
Then as now, whenever one speaks of freedom, one tends to think of it negatively, in terms of an absence offeree; when freedom is defined posi- tively, one thinks of it as oriented toward one's own (yet universally valid) reason. Since the positive determination of freedom is subject to semantic corrosion, only the negative definition remained stable and continues to be propagated today (on the basis of varying notions offeree) by liberal and socialist ideologues alike.
Following Kant, Schiller proclaims "that the laws by which the mind proceeds are not represented, and since they meet no resistance, they do not
53
(This does not exclude the possibility that a single artwork might be
the necessity that manifests itself in the artwork, one also sought evidence
to the contrary. Thus Schiller writes that "the imagination, even in its free
54
This presupposes a cognitive notion of freedom that, under frame conditions it must accept, makes room for
55
possible choices.
capabilities and one's imagination--generates the freedom to make deci- sions on the basis of which one can continue one's work. The freedoms
play, orients itselftoward boundaries. "
This version of the problem regis-
In this sense, creating a work of art--according to one's
204 Self-Organization: Coding and Programming
and necessities one encounters are entirely the products of art itself; they are consequences of decisions made within the work. The "necessity" of certain consequences one experiences in one's work or in the encounter with an artwork is not imposed by laws but results from the fact that one began, and how. This entails the risk of running into "insoluble prob- lems," problems that could never arise on the basis of laws.
The concept of self-programming solves the problems associated with
the traditional notion of freedom by relating freedom to self-generated
cognitive models. Self-programming does not mean that the individual
work is an autopoietic, self-generating system. But one can say that the
work constitutes the conditions of possibility for its own decisions, that it
observes itself, or, more accurately, that it can be observed only as a self-
56
observer.
nize how the rules that govern the work's own formal decisions are derived from these decisions. It is not clear how to specify such propositions at the operative level. It might suffice to point out that the work delimits the ob- serving operations of any observer (producer or beholder) by suggesting which ones are feasible and successful, and which ones are impossible, ob- structive, or in need of correction.
The concept of self-programming is incompatible with the notion that
57
one can get at the work's "essence" by disregarding the "nonessential. " This notion assumed that there is such a thing as a distinguishable es- sence, a remainder, so to speak. Today, this proposition convinces hardly anyone; at best, it gives rise to diverging views about the essence of art and of the artwork. The notion of omission confounds the structural levels of coding and programming. The positive/negative distinction must be exe- cuted in applying the binary code to all cases. Without this distinction, nothing comes about. But what is accepted or excluded by the system can be decided only on the basis of a program. The "essence" of art is the self- programming of the artwork.
If the classical formulation emphasizes freedom, then this means that there are no rules or concepts to guide a "critical," cognitive apprehension of beauty. But it also follows that art must operate without prohibiting the opposite. In deciding what fits and what doesn't, art follows patterns al- ready established by the work; it cannot rely on independently existing criteria that would prescribe what is permitted and what is not. This seemingly gives rise to a situation in which art can orient itself--and in this sense, art becomes historical--only in its own history, the history of
In order to observe a work of art adequately, one must recog-
Self-Organization: Coding and Programming 2. 05
an individual work's production and reception, be it the history of styles, or the intertextuality of the art system itself.
But can self-programming still be programming, if this concept applies, as it usually does, to the conditioning of something other? What would be the identity of the "self" that is the object of its own programming? Fur- thermore, how can the self-programming artwork distinguish itself, if it is no longer distinguished from the inaccessible domain it symbolizes or from the object it signifies by imitation?
These problems were confronted for the first time in the romantic re- flection on art. The distinction that guides this reflection was now located entirely within the art system. The individual work of art identifies itself by its distance from the Idea of art, which the work reflects upon in the impossibility of attaining it. Every work of art must be art as such, art in general; the romantic notion of "art criticism" insists on that. But the Idea remains an idea, whereas the artwork must be concrete. It must be acces- sible to the senses, and yet it must transcend itself. What is "characteris- tic" in the artwork can therefore not be apprehended in sensuous experi- ence, nor can it be reduced to the effect of a cause. It aligns itself with the Idea of art without being able to represent that Idea. And the form for this is self-programming--the notion that the work gives itself a form and thereby determines what is possible in the work and what is excluded. In order to express this notion, romanticism has recourse to the symbol and,
58
in so doing, goes beyond Kant.
This means only that difference--in this case the work's distance from
the Idea--is understood as a unity. Contrary to the religious tradition, dis- tinction and unity are now located entirely within the autonomous realm of art, reflecting the fact that art has become autonomous. If this paradox is now called "self-programming," does that mean it has been explained?
Let us go further: self-programming is a case of self-reference. Self- reference can be practiced only if it can distinguish what it refers to. It pre- supposes the distinction between self-reference and hetero-reference.