No More Learning

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. Thus one arrives at the question: What is the hetero-reference of the self- programming artwork?
According to the logic of second-order observation, the hetero-reference of art can only be that which is rendered invisible by the distinguishing schema of first-order observation (as an observation of an observer, second- order observation is also an observation of the first order). Accordingly, het- ero-reference refers to that which is rendered invisible when distinctions are
206 Self-Organization: CodingandProgramming
introduced into die world: it refers to the irreducible unity of the world as the always operative unmarked space. The program guarantees that the work of art can establish itself at the level of second-order observation, no matter in what concrete form. Or, put differently, it secures the program dependency, the contingency of all the operations that produce or en- counter the artwork in a world that as a world cannot be contingent, a world that makes possible the introduction of distinctions for its own ob- servation by withdrawing, as a world, from observation (distinction). In this way, the program prevents the collapse of two distinctions that must re- main separate, namely, the distinction between self-reference and hetero- reference, on the one hand, and between the positive and negative values of
59
the code, on the other --for it goes without saying that the artwork can-
not think of itself as a success and consider the world a failure.
This conception excludes the notion of the world (or society) as a source of directives for the execution of artworks. We have called this ex- clusion the autonomy of the art system while making the sociological as- sumption that world autonomy can be accomplished only via societal au- tonomy. This means that the directives for elaborating and evaluating a
work of art must be derived from the work itself.
In many cases, the observational possibilities provided by the artwork
can be rendered visible by means of persons: as, for example, in a painting
that incorporates unified perspective; in a building that at once offers and
denies observational possibilities to those inside and outside the building;
in drama, which stages the difference between seeing (knowing) and not
seeing (ignorance) in front of an audience; or in the novel, which does the
same for the reader. This strategy can be emphasized and brought to an
unbeatable conclusion when a play is staged within a play (or, more sim-
ply, through lying and deception), or when a novel illustrates how Don
Quixote or Emma Bovary creates his or her own destiny through a self-
60
inspiring reading.
The availability of an unambiguous metaperspective that is related to
persons and reflects on romanticism must have been responsible for mak- ing poetry the paradigm of art in general. But this claim becomes unten- able once the observer is conceived in correspondingly abstract terms (as we are doing here) and is defined as an application of distinctions for the indication of one, rather than the other side of a form. The artwork can then be understood as a frame for the observation of observational possi- bilities that are included or excluded whenever someone uses a distinction.
The world of dihairesis--a world that used to be collectively accessible
Self-Organization: Coding and Programming 207
on the basis of given divisions--must be sacrificed. At the level of first-or- der observation, there might still be error, dishonesty, deception, machi- natio, and so forth, all of which can and must be corrected at this level. At the level of second-order observation, however, there are no longer any di- visions, there are only distinctions. The problem no longer resides in the potential need for correction; rather, the problem is that observation re- mains invisible to itself, no matter what. The self-programming of art is the form of expressing that this is so and that the world, as the condition for introducing distinctions, remains invisible--no matter what kind of operative injunctions are issued by the program.
All of this has consequences for the relationship between program and operation. A first-order observer who has begun working on, or looking at, a work of art (without a beginning there would be nothing to observe) can build on what is already there and search for what fits or might not fit. To such an observer, freedom appears as the constraint on further options. As a second-order observer, he can try to find out whether, and in what ways, other observers might observe his formal decisions. This accounts for the chronic feeling of "being misunderstood. " For what could guaran- tee that multiple observers read the same formative freedom into any given object? An observer of the third order, one in search of theoretical formulations, can only establish the existence of circular relationships. A program is the result of the operations it programs. Nothing else is meant by "self-programming. " At the same time, however, the second-order ob- server can see that the first-order observer might see things differently, so xhatfor neither of them does the tautology turn into paradox, and both can tell how it is possible to determine the next step.
IV
The notion of the self-programming artwork remains unsatisfactory in at least one respect. It raises the question of whether one should think of artworks as completely isolated from one another, or whether the pro- gramming must be programmed on its part--a notion that amounts to returning, albeit in a different form, to something like a rule-based art. Perhaps it was this unresolved question that prevented the individual work from being released completely into autonomy. Wouldn't one have to conclude that art emerges from chance or, at least, assume a new be- ginning in each individual case?
Evidence to the contrary was promptly found--in the realm of histori-
208 Self-Organization: CodingandProgramming
cal empiricism, so to speak. While observing broader contexts, one dis- covered that works of art influence the emergence of other works, even when imitation is prohibited. Winckelmann was perhaps the first to ex-
61
ploit this insight for a historiography of art ordered in periods. toricization of the self-description of the art system requires a periodization of art history (and vice versa). This move relates the concept of style to distinctions between periods; that is, it temporalizes a concept that had been familiar for a long time and initially referred to something like a type of design (maniera) or to genres of such types (like the curial style of text production or the stilo in rhetoric. ) This is not to say that a certain period is restricted to one particular style, nor does it mean that works of art lose their value when a certain style goes out of fashion. The acknowledgment of a plurality of styles interrupts the relationship between style and social class. The wealth of styles is now open to all observers who, as visitors of exhibitions or museums, are interested in art. From the viewpoint of the observer, inclusion in the art system renders style independent of prior
62
(social) stratification
might well disclose a correlation between the two, though this concerns only the interest in art rather than the preference for certain styles).
imposed by style.
(even though statistics, invisible in everyday life,
The form of style does not affect the autonomy of the artwork; it merely keeps stylistic deviation in check or permits such deviation (if the deviation succeeds). In this way, the canonization of style might encour- age the transition to another style, that is, it might stimulate evolution--
63
"defining itself and then escaping from its own definition. "
different kinds of style supersede one another, one can see--at a macro- level, so to speak--that art produces novelty and how, then, after ex- hausting the possibilities of a given style, moves on to the next. One can then recommend stylistic purity, recognize mixed stylistic forms, or regis-
64
ter their existence with astonishment.
style of mixing styles as a measure against stylistic purity.
There is an obvious temptation to make works of art converse with one another via their stylistic forms, that is, in view of the limited tolerance
65
This is why a junctional definition of style suggests it- self, one that can respond to the problem of how diverse artworks form a
66
network, so that a system of art can establish itself. By and large, this
functional definition covers what has been understood historically as "style"--both in the traditional sense of a design type and in the modern sense of historical styles that have a time of their own and become obso-
One might even recommend the
The his-
Observing
Self-Organization: Coding and Programming 209
lete once it is over. The paradigmatic significance of individual artworks that served as models to be copied fulfilled the same function, that is, it served as a functional equivalent to style. At the same time, the relation- ship of functional equivalence indicates that the increasing emphasis on originality--if not on the uniqueness of "authentic works," including the critique of copying--abandons style to this functional realm and encour- ages the observation of particularly impressive works of art in terms of style. If works of art are not allowed to be copies and have no style either, then they lose their significance as artworks. Singularities resist classifica- tion and therefore cannot be understood and observed as art. The act of assigning a work to a certain style signals that the work belongs to art. That is to say, there is a program-related possibility for representing art within the artwork, which exists apart from coding.
Does this amount to metaprogramming? Can one expect or demand that the artist search for and identify a style with which he then aligns his work? And has the classification of styles become indispensable to a com- petent art criticism?
It is doubtful that such notions can be sustained. The discussion of style, which by the nineteenth century had made a vain effort to clarify its self-understanding, is a clear indication. One sought the programmatic
67
only to use it for restorative purposes.
notion--especially if the motive is to find one's own style--that style is a matter of applying prefabricated formal decisions that owe their emer- gence to a work-dependent sense of what is fitting. To the observer and copier of styles, a style presents itself as a synopsis stabilized by habit, while he is aware that this is the side-effect of a spontaneous, merely code- oriented practice that has abandoned itself to the self-programming of the artwork. To emphasize this, one speaks of spontaneity or of an uncon- scious genesis of style, but spontaneity cannot be expected to occur twice. Choosing familiar styles as programs in an easily recognizable manner amounts to making a rather cheap claim to belong to the art system, and
68
often the works end up not being very convincing.
such degenerative trends are temporally marked as "new" (new gothic, and so on) or, when there is too much of that, as "post" (postmodernism). This seems to say that the demand for novelty, and thus for creativity, is constrained neither by style nor by the imitation of style. Any work of art can search for an as yet unoccupied niche within the context of family re- semblances between styles; it can probe new, "impressionistic" light con-
Yet it is difficult to appreciate the
Not accidentally,
210 Self-Organization: Coding and Programming
ditions in fields and forests, in cathedrals or train stations. It can also ex- press itself by rebelling against the constraints of style. The classification of styles can be left to the art expert who, like a botanist, consults hand- books on stylistics in order to determine a given style. In this respect, style is not a program but a formal model with which or against which one can work. The outward boundary against the unmarked space of the world is shifted once again. The task of rendering visible the invisible is distributed among two authorities, whose collaboration covers up the fact that this is at stake.
Accordingly, the concept of style is a concept of difference, that is, a concept of form. The limitations of a style yield the possibility for further styles--although at first only in the unmarked space of world possibilities. Historically, just this possibility offers the temptation to make the transi- tion to a new style. The test is that the work must succeed as an artwork. In this manner, one gains the impression of a plurality of styles that have proven themselves in artworks as if by evolutionary selection. This makes possible the ultimate form of reflection--the "postmodern" style of mix- ing styles--in which the sovereign self-programming of the artwork can be displayed once again. But the combination of diverse stylistic quota- tions as such does not yet constitute a program. It might succeed or fail. It must confront the code of art. Otherwise it will not be recognized as art.
? 6 Evolution
I
We know a great deal about the history of art. Ever since the forms and artworks inherited by tradition lost their binding force and ceased to serve as models--that is to say, since the eighteenth century--the historiogra- phy of art has amassed a tremendous amount of knowledge. And ever since one began to compare artifacts in an historically and regionally far- reaching manner, there has been "culture"--no longer in the sense of "cul- tivation o f . . . " (of agriculture or cultura animi), but as an elevated sphere of reality, a level at which all testimonies to human activity are registered a second time--not with an eye toward their utilitarian significance but in comparison to other cultural testimonies. In comparison, works of art
(but also religions, institutions of law, or forms of social order) appear "in- teresting," all the more so, the more the comparison ventures into realms that are foreign, bizarre, strange, or difficult to comprehend. Considered as culture, art and religion appear to be universals of human society, but only on the basis of the specifically European and specifically historical point of view that is interested in such comparisons and constructs com- parative viewpoints. As a result, one now finds art in places where neither the producer nor the viewer knew that art was at stake, let alone culture. This difference is reflected upon, for example, in Schiller's distinction be- tween naive and sentimental poetry.
With art, just as with religion, the observation of culture--a kind of second-order observation--must have had disastrous consequences. In or- der to compensate for that, culture is emphatically affirmed and cele-
211
2 1 2 Evolution
brated as a value sphere of a special kind. At die same time, culture suffers from a broken heart, reflecting upon its own reflection and registering what has been lost and will never be created again in terms of nai'vete\ Ob- serving works of art under such circumstances requires blinders that shut out culture, arid yet, their benefit is dubious when the works are already infected by culture, when they are produced with an eye to comparison and, as a result, cannot be adequately understood in a naive mode. Or can they? Could it be that including the exclusion of comparative culture has
1 become an essential component in the observation of art?
The differentiation of an academic art history seems to respond to pre-
cisely this problem by offering the possibility for distinguishing between
an observation as art and an observation as culture. Art historical knowl-
edge in part interprets individual works or masters within their own tem-
poral and historical horizons, in part reconstructs relationships of influence
by tracing presumed causalities. Art historians also analyze developmental
trends, whether or not they presume a historical progress of some kind. An
academic discipline established exclusively for this task has existed only for
2
about a century. For the collection and proliferation of such knowledge,
3
"sources" are significant. These "mouse-eaten records" count only when
they appear authentic in the eyes of art-historical knowledge. Authenticity almost suffices to legitimize a source as remarkable. When studying Vero- nese, one cannot afford to disregard a single work by this painter. Veronese is Veronese.
Following Dilthey, one tends to think that the historian's task is to ren- der totalities in the form of individual figures and to contextualize details. Such a task justifies a selective use of sources, above all a disregard for what happened later and could not have been known when the work was conceived. Of course, the historian has an obligation to investigate the past that was known at the time the artworks of interest to him were cre- ated. This explains the inclination (or compulsion? ) to see the totalities constructed by the humanities as historical totalities, whose temporal horizons vanished with them but can be rediscovered in the present as our own past. In this way, historiography, as well as the historiography of art, combines the binding force of origins with a (henceforth only) historical relevance. Both disciplines present temporal figures within a reflexive tem- poral horizon--our own--that modifies temporal horizons in time and through time. In addition, one discovers everyday worlds against which high cultures stand out as esoteric exceptions; or one can demonstrate, by
Evolution 213
means of quantitative or statistical analyses, "latent structures" that show how knowledge swims in an ocean of ignorance.
All of this is well known and, being current knowledge, it suggests a tempting proximity to our own concerns. What is remarkable forces itself upon us. This makes it all the more necessary to preface our analysis with a clarification: an evolutionary theoretical analysis of history pursues goals of a different sort, and it orders its material in different ways. It rests on a specific theoretical formulation of a problem. In biology, this problem can be stated as follows: How can the one-time biochemical invention of self- reproducing life give rise to such a great diversity of species? For the the- ory of society, the problem is how to account for the high degree of struc- tural complexity that develops once a continuous, rather than sporadic and repeatedly interrupted communication has been secured--where the structural complexity in question might concern a multiplicity of histori- cal societies or the modern world society. In Spencers well-known formu- lation, this implies a "change from a state of indefinite, incoherent homo-
4
geneity to a state of definite, coherent homogeneity. "
about the system of society is the diversity of functional systems, and within these systems, the emergence of media that facilitate rich, if unsta- ble, formations--ever new transactions in the economy together with the systems of production that accompany such transactions, or the continual modification of a positive law that remains nonetheless stable. A theoret- ical interest that goes under the name of evolutionary theory focuses on the conditions of possibility for structural change and, constrained by this focus, on explaining the emergence of structural and semantic complex- ity. This implies that the description of art, the emergence of a new con- cept of culture, the cultivation of art as culture, and even the emergence of a theory of evolution must be understood as a result of evolution. The theory of evolution is a self-referential, an "autological" paradigm.
The scientific usage of the term evolution is not necessarily so precise. Especially in the social sciences, pre-Darwinist notions prevail. Merely de- scriptive phase models of social development--in vogue since the eigh- teenth century (thus long before Compte)--are frequently offered as a theory of evolution. There might be an explanation for this. One might argue, for example, that "social Darwinism" has never been persuasive in the social sciences, that one needs dynamic models capable of explaining why things today are no longer the way they used to be, or that, while evolutionary adaptation to accidental structural changes cannot be de-
What is impressive
214 Evolution
nied, such changes might be analyzed more adequately with reference to
5
Lamarck rather than to Darwin. In a rigorous and precise conceptual
sense, none of these approaches deserves to be called a theory of evolu-
tion. This is why the theory of evolution has been called an "untried the-
6
ory," and justifiably so. This is all right--at least according to the many
social scientists who reject evolutionary theory as a biological metaphor or as an illegitimate analogy to the world of organisms.
Specifying a line of inquiry that might be called evolutionary theory (but could certainly assume other names as well) is an indispensable pre- paratory step, even though it says little about the research program. The theory of evolution deploys a specific distinction, namely, the distinction between variety, selection, and restabilization. This line of questioning does not focus on a process, nor does it attempt to explain in a historical or causal manner why things happen the way they do. Rather, it is moti- vated by systems-theoretical concerns. If autopoietic systems are set up in such a way that they must use their own operations to create and modify, or forget and dispose of their own structures, and if this mode of operat- ing always presupposes a potential network of operations--that is, a structure--then the question arises: How is it possible that these struc- tures become increasingly complex?
Above all, this tendency is unlikely. What makes it more likely? And how does improbability--the fact that certain phrases are uttered nevertheless, that certain goods are bought, and certain forms are created and admired as art--eventually become so plau- sible that one can almost count on it? How can society establish its own improbabilities (which require that something specific must be selected from coundess other possibilities) in such a way that they stabilize one an- other, and the sudden failure of crucial accomplishments (for example, the loss of the monetary economy or the police force) would result in a catastrophe whose consequences would be impossible to contain? How, in other words, is it possible that the improbability of emergence continually
7
transforms itself into the probability of preservation?
The theory of evolution is concerned with unfolding a paradox, namely,
the paradoxical probability of the improbable. However, we cannot help formulating this paradox in a manner that statisticians will not accept. In statistics, it is trivial that reality, in each of its expressions, is extremely un- likely and at the same time entirely normal. It is therefore not surprising that the statistician fails to register this paradox, precisely because he pre- supposes its unfolding. The same holds for the theory of evolution. The
Evolution 215
comparison illustrates, however, that recourse to paradox--no matter how insignificant its methodological benefits might be, and inasmuch as it must be prohibited methodologically--allows one to raise the theoretical question of what kinds of identification facilitate, in one way or another, the unfolding (= rendering invisible) of the paradox. This paradox is ulti- mately a paradox of self-implication, which resides in presupposing a dis- tinction (here, probable/improbable) whose unity can be indicated only paradoxically. Logicians might object that theory creates this puzzle only to solve it on its own. This is certainly true. The question is: What kinds of comparative possibilities become visible in this manner?
II
One can present the history of society as the history of a general socio-
8
cultural evolution.
system of society at large. Changes in the realm of art would appear only as instances in the evolution of society. As early as 1800, this problem was discussed--albeit without sufficient theoretical preparation--with refer- ence to Kant's legal-political concept of society and in view of the rising
9 expectations being placed upon art and aesthetic experience. If one as-
sumes an elaborated theory of evolution instead of a theory of conscious- ness, then the question becomes whether independent (albeit condi- tioned) subsystems can exist within evolving systems. In order to prove this assumption, one would have to show how, and under what condi- tions, autopoietic subsystems close themselves off and, by differentiating operational modes of their own, become capable of treating environmen- tal perturbations as chance events that stimulate the variation and selec- tion of system-internal structures.
We have addressed this issue in conjunction with the historical condi-
10
tions of the art system's differentiation.
to furnish evidence for special environmental conditions that favored dif- ferentiation. In the following, we seek to identify the evolutionary mech- anisms whose separation facilitates this process.
Let us begin by recapitulating our analyses of the artworks form. Already in the individual artwork, we can see how the improbability of emergence is transformed into the probability of preservation. The first distinction, the one from which the artist starts out, cannot be programmed by the work of art. It can only occur spontaneously--even though it implies a decision
But the system reference of such a history remains the
In that context, our concern was
2 l 6 Evolution
concerning the work's type (whether it is to be a poem, a fugue, or a glass window) and perhaps an idea in the artist's mind. Any further decision tightens the work, orienting itself toward what is already there, specifying the unoccupied sides of already established forms and restricting the free- dom of further decisions. Once the distinctions begin to stabilize and relate to one another recursively, what occurs is precisely what we expect from evolution: the artwork finds stability within itself; it can be recognized and observed repeatedly. The work might still suffer destruction, but any fur- ther modification becomes increasingly difficult. Some insoluble problems or imperfections might remain, which must be accepted as a matter of fact. Even in art, evolution does not bring about perfect conditions.
A work might also be conceived more or less according to plan. As in politics or in the economy, the plan becomes a part of evolution. If the artist adheres rigidly to a preconceived program, then he will either pro- duce works devoid of qualitative differences (even if he applies different programs), or he will have to decide between simple acceptance or rejec- tion of the work as a whole. Typically, however, the artist allows himself to be irritated and informed by the emerging work, whatever the program might entail. The typical case is evolution.
It is perhaps a unique feature of the art system that the "intertextual" network connecting works produced within the system is not very tight, and that, to use a strong formulation, chance events are already trans- formed into necessities at this level. When searching for a theory of how the art system evolves, we must keep this small-scale revolution of the in- dividual work in mind. But the evolutionary mechanisms of variety, se- lection, and restabilization differentiate themselves only at the level of the system. Only at this level do social conditions emerge that facilitate the production of artworks. If art is not sufficiendy differentiated as a phe- nomenon, then there can be no freedom of beginning, no conception of what is involved in producing or encountering a work of art.
The theory of form combination, which we take as our starting point,
suggests that art originates in the ornament, under conditions that imply
no awareness of a corresponding concept, let alone of an autonomous art
11
system.
nament is to the evolving art system what the evolution of language is to the evolution of society; in both cases, there is an extended preparatory stage that yields eruptive consequences once communication has been sta- bilized to the point where its boundaries become visible. What is empha-
One might propose a bold comparison: the evolution of the or-
Evolution 2 1 7
sized at first, however, is not the difference between object and adorn- ment, but rather the unity of this difference, its meaning. "Cosmos" in the Greek sense means both order and ornament.
In prehistoric times, ornaments emerged independently everywhere in the world (even though the question of whether some patterns emerged independently or by diffusion is a matter of dispute). In premodern soci- eties, the relationship between surface and depth was experienced differ- ently from how it is today. This is evident in the widely used techniques of divination. These techniques are concerned with displaying signs on a vis- ible surface, signs that betray depth. Perhaps ornaments were understood in the same way.
The ornament provided an opportunity to train oneself artistically without depending on demanding social presuppositions. The basis for this trend was a well-developed competence in technique and skill that might have produced ornamental order as a side effect, as a playful addi- tion superimposed on something useful and necessary--as adornment. One could follow the inspiration of existing models or use as guidance the limitations of cult objects or other objects of utility. In this way, one could profit from the integration of such objects in nonartistic contexts and from their evolutionary differentiation. The ornament provided an op- portunity to practice observation and to train one's eyes and hands for a type of social communication that could later exploit such skills to create a self-differentiating system.
There might be enough material stashed away somewhere in libraries to write a history of the ornament that could tell what kinds of figurative patterns were used to decorate objects: some ornaments are geometrical, whereas others move in waving lines; some patterns display protruding, recognizable leaves, fruits, heads, and so on; some ornaments are stacked atop each other; whereas others support the formal play of the objects they decorate--a vase, an oven grid, a door, a building--whether for em- phasis or to cover up imperfections, whether to make believe or to join
12
figures. Perhaps there are such compilations,
ory of art, they would serve at best as illustrative materials that one might also find elsewhere.
Distinguishing between a historical account and a theory of evolution is imperative. The prime concern of a theory of evolution is to account for discontinuities and structural changes that suddenly erupt after extended periods of stagnation or incremental growth. Such a theory focuses on the
but for an evolutionary the-
2 l 8 Evolution
prolonged irritation to which forms are exposed and, above all, on the abrupt occurrence of operative closure, with its chances for autopoietic autonomy.
From this viewpoint, the practice of decoration (in the widest sense) appears to be a preadaptive advance, a development that initially served other functions and to which one can return in the course of the art sys- tem's differentiation as if art had existed at all times. Once a system of art begins to differentiate itself, it becomes possible to construct a past; one can redirect a treasure house of forms and continue to use skills one al- ready has. In this way, a structural break in the social domain is at first ex- perienced only as an artistic innovation, as an improvement in skill. Un- der radically new social conditions, art initially sought less radical forms of expression--one returned to antiquity, began to valorize the artists so- cial prestige, and sought independence from the directives of patrons-- and only gradually were novelty and originality demanded from the indi- vidual artwork.
The differentiation of the art system must have altered the meaning of ornamentation, in particular by adding a dimension of "depth," so that today only the combination of forms as such is important. In gothic ar- chitecture, the ornament was already taken in tow by inventions in struc- tural form, within which it had to prove itself. Subsequent reflections on the limitations of ornamentation and on the primacy of balanced pro- portion could draw on a history that rendered such developments plausi- ble. The distinction between form and supporting decoration could thus be generalized and adopted later as a theory of the self-differentiating art system. With the emergence of self-conscious artworks that insist on be- ing recognized as such, the traditional domain of artistic skill was divided into two separate realms: one in which decorating objects of utility pre- vails and where one later turned to certain "crafts" to compete against in- dustrial production; and a realm of art in which works must decide for themselves whether they need or can tolerate ornamentation, and if so, to what extent and in what form. At first, divisions of this kind were neces-
13 sary. Following Alberti--who introduced the notion of composition --
the standard Renaissance literature distinguishes between drawing, com-
14
position, and coloration as necessary components of painting. concept of the drawing, of contour or design, continues the tradition of
15
the ornament in a form reduced to one of its components.
cento in Italy, especially in Florence, developed a theory of disegno that
The cinque-
The
Evolution 2 1 9
covers the entire problematic, stretching the concept to the point where
16
it loses its precision.
conception (and in this respect, it resembles God's creation of the world, that is, nature in its entirety), while on the other hand it also indicates the artful execution of the work by skilled eyes and hands. Disegno involves invention, ingenuity, and intellect (in the traditional sense), while being concerned with a technique of signs, with skills taught in academies, and with the form and contours of the work itself. Since this contradiction could not be resolved, the seventeenth-century discussion of the concept ran out of steam, leaving behind a theory of drawing skills that could be taught.
Poetry follows similar distinctions. Torquato Tasso, for example, divides
17
his Discorsi dell'artepoetica e in particolare sopra ilpoema eroico into ma-
teria, forma, and ornamenti, only to focus entirely on materia (choice of
18
topic) and forma. In his treatment of ornamenti Tasso changes his tone, speaking of elocuzione and remaining entirely within the framework of rhetorical distinctions of style, which might equally well be treated as dis- tinctions of form.
Parallel to this discussion, one finds--now under the heading of the or- nament--a degradation of the ornamental to mere decoration or adorn- ment. The low esteem in which the ornament was held raises the issue of whether works of art require ornamentation, and if so, why. The solution was to relegate the ornament to a subordinate, merely decorative function in every realm, including the arts, and to distinguish its supplementary
19
function from a more important type of beauty in both nature and art. In this way, one could playfully adjust, at the level of ornamentation, to social changes and develop or adopt forms that did not interfere with the work's thematic focus. One could turn away from a merely religious sym- bolism and influence the development of styles by drawing on natural
20
forms, interpersonal relationships, heraldry, or models from antiquity. But the distinction between art and ornament (whether in the work of art itself or in other objects) undermines the possibility for indicating the unity of art; if beauty--understood as perfection--requires a supplement,
21
it is not clear what is meant by unity.
master/slave metaphor loses plausibility; besides, the focus shifts toward what holds the work of art together from within. Only linguistic usage stands in the way of responding spontaneously: the ornament.
The distinction, introduced by Hutcheson, between original (or abso-
On the one hand, disegno stands for the creative
In the eighteenth century, this
2 2 0 Evolution
22
lute) and comparative (or relative) beauty should not be underestimated. It is a decisive step toward rehabilitating the ornament and toward push-
23
ing back the semantics of imitation.
Original or absolute beauty is noth-
ing other than (the subjective idea of) the ornamental. Hutcheson defines
this type of beauty ("to speak in the mathematical style") in terms of a
"uniformity amidst variety" or as a "compound ratio of uniformity and va-
24
riety. " Since this formula--reminiscent of Leibniz--embraces too much
(according to Leibniz, it comprehends the entire world), one introduces a principle of intensification that renders a given variety more uniform or adds variety to uniformity. According to the associational psychology popular at the time, even failure and ugliness can be calculated in the
25 form of a disruption brought about by associations that do not fit.
The framing of this concept in epistemology and moral theory (psy- chology) is limited historically, and a philosophical aesthetics will pursue different goals. But the effects of ornamentation and continued references to the ornament are noticeable, especially in William Hogarth. In his es- say The Analysis of Beauty, Hogarth still mentions the ornament, but goes on to describe the movement of line as a principle of enhancement, which culminates in "serpentine lines" that present the "inner surface" of the ob- ject and its potential movement in its most favorable proportions. This in- sight into the function of drawing can be translated into technical in- structions for producing beauty that are intelligible to anyone (not only to "connoisseurs," who operate according to obscure principles), thus ac-
26
complishing a comprehensive inclusion of observers into the work of art. By and large, the long tradition of statements on the line in drawing re- mains ambivalent. On the one hand, such statements remain subordinate to an interest in beauty, harmony, and balanced proportion; on the other hand, they gain significance to the extent that the vacuity and redundancy of such a notion of beauty becomes apparent. This is evident not only in
27
Hogarth but also in Moritz and Herder.
To the extent that problems of form acquire a dimension of depth and
one began to turn what one had learned from the ornament (for example, under the name of disegno) into a theory of the artwork, there are tenden- cies to rescue the ornament in its exorbitant, if not to say superfluous function, to reinstate it as a kind of supplement, or to use it as a way of transcending the perfection one strives for. This happened in mannerism, which legitimizes capricious and fantastic trends that explode the limits of proportion. Zuccaro presents a theoretical integration of this possibility
that explicitly refers to the ornament.
28
The two forms of disegno, which
Evolution 2 2 1
combine imitation and perfection, are supplemented by a third--the
bizarre, capricious disegnofantastico--which adds variety (diversitO) to the
29
already perfect artwork.
Classicizing theory in the second half of the eighteenth century dealt
extensively with the ornament (adornment, arabesque) in the hope of bal- ancing the sterility of forms, on the one hand, and die lack of discipline, on the other, and in order to test the classicist idea of style in the subordi-
30
nate realm of decoration.
uncontrolled excess of arabesques and grotesques and their proximity to chaos calls attention to itself, as if the problem of disorder that underlies
31
In the transition to romanticism, precisely the
all creative activity could be harnessed in this undomesticated form. internal dynamic involved in the rehabilitation of trie ornament has been
32
investigated especially by Gustav Rene Hocke.
If one searches for an analogue to the intensification of the ornamental
outside of the visual arts, then one is likely to discover that suspense fulfills
33
a similar function of intensifying narrative structure in literature.
thematic level, the demand that the narration be charged with suspense leads to the disengagement of the hero from the effects of an external fate--a fate that, in early modern times, had functioned as a useful device
34
for increasing variety within the framework of typified redundancies. The narrative development of characters interrupts the nexus between past and future. One needs actions in order to establish coherence, and actions require motives; only toward the end of the story does it become apparent why things happened the way they did. Narrative moves its plot as if in serpentine lines; it fills a space of self-generated uncertainty, so that, in the end, the meaning of the plot can enter the plot (the couple gets married, the criminal is recognized and punished). The narrative--or the play, to speak with Dryden--must be constructed like a labyrinth, in which the spectator can see just a few steps ahead and the conclusion is not recog-
35
nized until the end.
draws variety into the work itself, which earlier had to be supplemented from the outside; this means that the author must know what the reader is not yet allowed to know. If suspense secures the work's unity (as an orna- ment does), then the characteristic features of persons can be rendered more individually without any loss of recognizability. The level of the work where forms are combined permits greater variety while preserving the redundancy necessary to generate information.
Suspense, in the sense of self-generated uncertainty,
The
On the
2 2 Z Evolution
What does all this have to do with the ornament? The ornament, too,
36
strives for a complex level of redundancy and variety
37
--in Hogarth's for-
mulation, for "the art of varying well" --as if in "serpentine lines. " Re-
dundancy is secured so long as the narrative contains enough detailed ref-
erences to the familiar world of the reader (short of serving him a story he
38
already knows! ).
ments are left open, but only a few of them concern the future (for the mystery novel, this means several possible ways of discovering the past). What is at stake in narrative, in other words, is the combination of con- nectivity and an open future. The question is which turn the line or the story might take. Prolonging the line under conditions of continued sus- pense amounts to crossing the boundary of the form while covering up
39
Suspense is created when several potential develop-
this crossing. It is therefore not surprising that Moritz,
the "metaphysical line of beauty" in the epic and in drama, emphasizes its strong curvature (in comparison with the line of truth) and what it omits, because it suggests the form of the closed circle. Nor should it strike us as odd that Friedrich Schlegel would call a novel (Diderot's Jacques lefatal- iste) an arabesque and object to the low esteem in which this form was held--according to Schlegel, the arabesque is "a fully determined and es-
40
sential form or mode of poetic expression. "
been suggested by Georg Lukacs, who claims that irony is the successor of
41
the ornament:
and downs of narrative events are played. We might call suspense, or per- haps irony, the inner forms of the novel's unity, forms that are compatible
42
In historical retrospect, an art produced in accordance with these prin-
ciples (and following these injunctions) might strike us as remarkable,
perhaps even as the culmination of European art. The second half of the
nineteenth century is preoccupied with the question of whether a careful
study of decorative style might be able to rejuvenate a style that was ev-
43
idently lacking.
again--one sacrificed the object in the visual arts, tonality in music, and the continuous story line in literature. By now, ornamentation had be- come what it has always been: a self-directing form combination, the tem- porality of observation, which is continually in search of what has yet to be decided.
But we still do not know how evolution managed to bring about this state of affairs.
irony is the persistently maintained key in which the ups
with, indeed demand, a great variety of narrated events.
Around 1900 the repertoire of styles expanded once
when speaking of
An alternative venue has
Evolution 223
III
The distinction with which the theory of evolution dissolves, displaces, represses, and renders invisible the paradoxical probability of the improb- able is the distinction between variation and selection--that is to say, an- otherdistinction. One can start all over again, if one can presuppose (which certainly does not go without saying) that variation and selection are sepa- rable in reality and can subsequently be distinguished by an observer.
In nineteenth-century theories of evolution, the notion of the "individ- ual" played a decisive role in explaining variation (as a precondition for se- lection). Of course, one needs to distinguish between two different ver- sions of the individual. Along with the concept of population, a collective individualism established itself against the traditional typological essen- tialism of the doctrine of species and genres. Populations are capable of evolution because they are made up of individuals. At first, one believes that the diversity of individual forms is the source of the adaptability of populations--that variety is a source of variation. Depending on the course of changing environmental conditions, one or the other character- istic grows stronger and is reproduced in large quantities. Applying this argument to human society, however, transforms it completely. Now, the large number of individuals increases the likelihood that some of them turn out to be creative, innovative, and powerful, and the statistical nor-
mality of such exceptional cases supports the explanation of evolutionary variation. No one would speak of particularly creative flies, birds, or frogs in order to account for changes in the behavior of a specific animal popu- lation. But for society, and especially for the realm of art, such explana- tions do make sense (at least ideologically), though it would be less plau- sible to focus on the diversity that exists in the form of a "population" of individual artists or works of art.
A long-standing cult of genius paved the way for the explanation of evo-
lution in terms of the individual. In retrospect, one can rephrase Kant's
distinction between genius (variety) and taste (selection) as a theory of
44
Thinking of variation and selection as internal functions of a
evolution.
system's evolution precludes the possibility of attributing the cause of evo-
45
lution or innovation to "great men and women. "
thinking in historical terms faced the problem of explaining why at certain times geniuses appear in large numbers while at other times they are nowhere to be found. Irregularities of this sort might be treated as a pecu-
An age that was already
2 2 4 Evolution
liarity of certain historical periods and charged to the account of the times themselves, which sometimes nourish and sometimes don't. It would be more fruitful, however, to invert the relationship between these variables and think of "genius" as the product rather than the cause of evolution. "Genius" stands for the improbability of emergence, and "taste" for the likelihood that works of art prevail. Genius must be admired; taste must be justified.
At first, this distinction appears as sheer difference, without a concept for the unity of what is distinguished. (This difference is accounted for, so to speak, by the creative power of genius). By means of a special trick, however, the theory of evolution can nonetheless come to terms with the unity of the distinction between variety and selection--namely, by posi- tioning this unity and the distinction side by side. The unity of the dis- tinction then assumes the name of a third, namely, stabilization or resta-
bilization. If there is variation--a positive or negative selection that takes into account or disregards a given variant in the reproduction of sys- tems--then it raises the questions: Under what kinds of structural condi- tions does the reproduction (of autopoietic systems) take place? How can a system continue to reproduce itself, if it accepts variation, or if it rejects
46
a possibility that offered itself (although other systems might use it )?
Problems of stabilization are not solely consequences of evolution; they do not solely occur after the fact. A system must already be stabilized if it is to offer opportunities for variation. Stability is the beginning and the end of evolution, a mode of structural change that simultaneously generates instability. This is why the evolutionary theoretical model, which abstracts from time, describes a circular relationship between variation, selection, and (re)stabilization. This is only an indication that the unfolding of the paradox takes time. It explains why, in superficial descriptions, evolution- ary theory is presented as a theory of processes. The systems-theoretical concept for this phenomenon is dynamic stability.
This abstract theoretical concept can be successfully applied to the em- pirical realm if one can show how in reality variation, selection, and (re)stabilization each depend on different conditions, in other words, if one can show that they occur in isolation. One tends to assume that the theory of evolution presupposes an accidental coordination of its mecha- nisms (rather than an integration that is contingent upon the system). The theory of organic evolution has successfully isolated these phenom- ena with concepts such as mutation, sexual reproduction, "natural selec-
Evolution 2 2 5
tion," or the selection of organisms for the reproduction and the ecologi- cal stabilization of populations. We need not concern ourselves with issues that are still debated within this (more or less "neo-Darwinian") theory, such as the notion of "adaptation" to the environment or "natural selec- tion. " At any rate, this entire apparatus for describing the functions of separation in biology is inapplicable in the domain of sociocultural or so- cial evolution. This is not to say that a theory of evolution cannot be for- mulated for society, but rather that functions of separation in this domain
47 must be described differently.
In systems theory, one can distinguish between operations (elements), structures, and the system, that is, one can discern a difference between sys- tem and environment. This distinction facilitates an appropriate attribu- tion of evolutionary mechanisms. One can speak of variation only where unexpected (new! ) operations occur. In these cases, selection concerns the structural value of an innovation: the innovation is either accepted as something worth repeating, or it is isolated as a singular occurrence and re- jected. Stability might be jeopardized in both cases, because new structures need to be integrated, and discarded innovations must be remembered or
48
perhaps become an object of regret. The sheer quantity of operations al-
lows trivial variations to occur on a gigantic scale, variations that, under
normal circumstances, vanish as soon as they take place. Occasionally their
structural value is recognized. In this case, selection becomes an issue.
When this happens, variation can endanger the system, exposing it to a
persistent pressure of irritation and forcing it to adapt internally to its own
49
problems.
This theoretical schema presupposes a system of sufficient complexity.
Evolutionary mechanisms cannot be thought of in isolation, unless one
can assume a "loose coupling" of multiple simultaneous operations, which
under normal circumstances ensures that variations are immediately elim-
inated; otherwise the pressure variation exerts upon structures would be
50
too high.
and tolerate structural change--in the sense of the older cybernetics, it must be organized in an "ultrastable" manner. Last but not least, evolution is possible only if the system can maintain the stability of prior and sub- sequent states and if it can distinguish between operations and structures, that is, between variations and selections. All of this precludes considering interactive systems among persons as capable of evolution, suggesting in- stead that the social system is the primary bearer of sociocultural evolu-
Apart from that, an evolving system must be able to localize
2 2 6 Evolution
tion. This raises the question--the only one of interest to us here-- whether one can speak of evolution in conjunction with social subsystems --specifically in conjunction with the art system.
Unlike the domain of evolutionary epistemology or the theory of sci- ence, in the domain of art hardly any preparatory work has been done for such analyses. In die past, evolutionary theories of social subdomains have typically been developed where, according to the self-understanding of the domain in question, problems of rationality have come to the fore: in sci- ence, for example, on the occasion of the transcendental-theoretical revo- lution and as a result of the current constructivist revolution; in the econ- omy because of doubts about whether the model of perfect competition can serve as a valid orientation; in law in view of the obsolescence of nat- ural law and the necessity of coming up with other (not just value-related) explanations for the selection of current law. It is evident that theories of evolution are also subject to evolution and that they tend to be advanced when doubts about rationality cannot be overcome in any other way. Art, however, has always thrived on the imagination, so that a typical occasion for evolutionary models of explanation never arose. Social-theoretical mod- els might conceivably be inadequate for applying the theory of evolution in die realm of art. Be this as it may, the nexus between systems theory and the theory of evolution outlined above could be an occasion to attempt an application of this sort with new theoretical tools.
IV
If one wants to apply the theoretical approach outlined above to art, one must first determine (just as in systems theory) the operation that provides the point of onset for variations. This must be the operation that supports whatever happens in art, which must not be confused with other opera- tions--otherwise one might end up with an evolution that has nothing to do with the system of art. Within the systems-theoretical framework we presented earlier, we can define this operation in only one way, namely, in terms of an observation that is focused on art. This notion covers both the production of art and the encounter with artworks. Formally, it indicates a specific way of choosing distinctions for the purpose of using one (but not the other) side as the starting point for further operations.