See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/290813605
Construction and Intuition: Creativity in Early Computer Art
Chapter · August 2012
DOI: 10.1007/978-3-642-31727-9_3
CITATIONS
6
READS
679
1 author:
Some of the authors of this publication are also working on these related projects:
The algorithmic dimension in art View project
Frieder Nake
Universität Bremen
80 PUBLICATIONS   231 CITATIONS   
SEE PROFILE
All content following this page was uploaded by Frieder Nake on 24 April 2017.
The user has requested enhancement of the downloaded file.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Chapter 3
Construction and Intuition: Creativity in Early
Computer Art
Frieder Nake
Abstract This chapter takes some facets from the early history of computer art (or
what would be better called “algorithmic art”), as the background for a discussion
of the question: how does the invention and use of algorithms influence creativity?
Marcel Duchamp’s position is positively referred to, according to which the spectator
and society play an important role in the creative process. If creativity is the
process of surmounting the resistance of some material, it is the algorithm that takes
on the role of the material in algorithmic art. Thus, creativity has become relative
to semiotic situations and processes more than to material situations and processes.
A small selection of works from the history of algorithmic art are used for case
studies.
3.1 Introduction
In the year 1998, the grand old man of German pedagogy, Hartmut von Hentig,
published a short essay on creativity. In less than seventy pages he discusses, as
the subtitle of his book announces, “high expectations of a weak concept” (Hentig
1998). He calls the concept of creativity “weak”. This could mean that it is not
leading far, it does not possess much expressive power, nor is it capable of drawing
a clear line. On the other hand, many may believe that creativity is a strong and
important concept.
Von Hentig’s treatise starts from the observation that epochs and cultures may
be characterised by great and powerful words. In their time, they became the call
to arms, the promise and aspiration that people would fight for. In ancient Greece,
Hentig suggests, those promises carried names like arete (excellence, living up to
one’s full potential), and agon (challenge in contest). In Rome this was fides (trust)
and pietas (devotion to duty), and in modern times this role went to humanitas,
enlightenment, progress, and performance. Hardly ever did an epoch truly live up
to what its great aspirations called for. But people’s activities and decisions, if only
ideologically, gained orientation from the bright light of the epoch’s promise.
F. Nake ( )
University of Bre Germany
e-mail: nake@informatik.uni-bremen.de
J. McCormack, M. d’Inverno (eds.), Computers and Creativity,
DOI 10.1007/978-3-642-31727-9_3, © Springer-Verlag Berlin Heidelberg 2012
61
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
62 F. Nake
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
If in current times we were in need of a single such concept, “creativity” would
probably be considered as one of the favourites. Information, communication, sustainability,
ecology, or globalisation might be competing. However, creativity would
probably still win. It is a concept full of shining promise. Nobody dares criticise it
as plastic and arbitrary. Everybody appears to be relating positively to it. Technofreaks
use it as well as environmentalists. No political party would drop it from their
rhetoric.
Creativity may be considered as a means for activity, or as its goal. However,
von Hentig is sceptical about the possibility of developing more creativity through
education and training; he is also sceptical about creative skills independent of the
context. Creativity as an abstract, general concept, taken out of context, is unlikely
to exist. If a helpful concept at all, creativity is bound to situations and contexts.
Only relative to them may our judgement evaluate an activity as creative. Creativity
exists only concretely.
Leaving out ancient Greece, the Middle Ages, and the Renaissance, it seems that
the way we understand “creativity” today is as a US-American invention (Hentig
1998, p. 12). It started with the fabulous definition of an IQ (Intelligence Quotient)
and operational tests to measure it by Stern (1912) in Germany. His approach became
an operational method in the USA by the end of World War I. J.P. Guilford
(1950) and others made clear that IQ tests did not identify anything that might be
called “creative”. Current creativity research starts from this article. Like any other
measure, a test of your IQ may at best say something about a standard behaviour
within given boundaries, but not much about crossing boundaries. Often people do
what they are supposed to do, and they do it well. Others do what they want to do,
and do it to the dismay of their bosses, teachers, or parents.
When we consider creativity as an attribute, a property, or a feature that we may
acquire by taking courses or joining training camps, we put creativity close to a
thing, or a commodity. We inadvertently transform a subjective activity or behaviour
into an objective thing. We may acquire many or few commodities, cheap or expensive
ones. But is quantity important for understanding creativity, or for becoming a
creative person? Doesn’t it make more sense to associate the term “creativity” with
behaviour, activity, situation, and context? The idea of attaching creativity to individuals
is probably what we are immediately inclined to think. But it may still not
be very helpful. Creativity seems to emerge in situations that involve several people,
who interact in different roles with favourable and unfavourable conditions and
events.1
We may align intelligence with making sense in a situation that makes sense.
If we do so, creativity could be viewed as making sense in situations of nonsense.
Dream and fantasy are, perhaps, more substantial to creative behaviour than anything
else.
1We are so much accustomed to thinking of creativity as an individual’s very special condition and
achievement that we react against a more communal and cooperative concept. It would, of course,
be foolish to assume individuals were not capable of creative acts. It would likewise be foolish to
assume they can do so without the work of others.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 63
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
With these introductory remarks I want to announce a sceptical distance to the
very concept of creativity. With only little doubt, a phenomenon seems to exist that
we find convenient to call by this name. A person engaged in a task that requires
a lot of work, imagination, endurance, meetings, walks, days and nights, music, or
only a flash in the mind, will use whatever means she can get hold of in pursuit of
her task. Even computers and the Internet may be helpful, and they, indeed, often
are. If the final result of such efforts is stamped as a “creative” product, is it then
sensible to ask the question: what software and other technical means contributed
to this creation? Not much, in my view. And certainly nothing that goes beyond
their instrumental character. More interesting is to study changes in the role of the
instrument as an instrument. The sorcerer’s broom is more than a broom only in
the eyes of the un-initiated. It is an expression of a human’s weakness, not of the
instrument’s clever strength.
Therefore, I find it hard to seriously discuss issues of the kind: how to enhance
creativity by computer? Or: how do our tools become creative? If anything is sure
about creativity, it is its nature as a quality. You cannot come by creativity in a
quantitative way, unless you reduce the concept to something trivial.
In this chapter, I will study a few examples of early computer art. The question
is: How did the use of computers influence creative work in the visual arts? The
very size and complexity of the computer, the division into hardware and software
must, at the time, have had a strong influence on artistic creativity. The approach will
be descriptive and discursive. I will not explain. Insight is with the reader and her
imagination, not with the black printed material. I will simply write and describe.
I cannot do much more.
The chapter is divided into four narrations. All four circle around processes of
art or, in a less loaded expression, around aesthetic objects and processes. The art
we will study here is, not surprisingly, algorithmically founded. It is done, as might
be said, by algorists.2 They are artists of a new kind: they think their works and let
machines carry them out. These artists live between aesthetics and algorithmics and,
insofar, they constitute a genuinely new species. They do art in postmodern times.
When they started in the 1960s, they were often called computer artists, a term
most of them hated. Meanwhile, their work is embraced by art history, they have
conquered a small sector of the art market, and their mode of working has become
ubiquitous.
The first narration will be about a kind of mathematical object. It is called a
polygon and it plays a very important role. The narration is also about randomness,
which at times is regarded as a machinic counterpart to creativity.
Three artists, Vera Molnar, Charles Csuri, and Manfred Mohr, will be the heroes
of the second narration. It will be on certain aspects of their work pertaining to our
general topic of creativity.
2There actually exists a group of artists who call themselves, “the algorists”. The group is only
loosely connected, they don’t build a group in the typical sense of artists’ groups that have existed
in the history of art. The term algorist may have been coined by Roman Verostko, or by JeanPierre
Hébert, or both. Manfred Mohr, Vera Molnar, Hans Dehlinger, Charles Csuri are some other
algorists.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
64 F. Nake
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
Two programs will be citizens first class in the third narration: Harold Cohen’s
AARON stands out as one of the most ambitious and successful artistic software
development projects of all time. It is an absolutely exceptional event in the art
world. Hardly known at all is a program Frieder Nake wrote in 1968/69. He boldly
called it Generative Art I. The two programs are creative productions, and they
were used for creative productions. Their approaches constitute opposite ends of a
spectrum.
The chapter comes to its close with a fourth narration: on creativity. The first
three ramblings lead up to this one. Is there a conclusion? There is a conclusion
insofar as it brings this chapter to a physical end. It is no conclusion insofar as our
stories cannot end. As Peter Lunenfeld has told us, digital media are caught in an
aesthetics of the unfinish (Lunenfeld 1999, p. 7). I like to say the same in different
words: the art in a work of digital art is to be found in the infinite class of works a
program may generate, and not in the individual pieces that only represent the class.
I must warn the reader, but only very gently. There may occasionally be a formula
from mathematics. Don’t give up when you see it. Rather read around it, if you like.
These creatures are as important as they are hard to understand, and they are as
beautiful as any piece of art. People say, Mona Lisa’s smile remains a riddle. What is
different, then, between this painting and a formula from probability theory? Please,
dear reader, enter postmodern times! We will be with you.
3.2 The First Narration: On Random Polygons
Polygons are often boringly simple figures when it comes to the generation of aesthetic,
or even artistic objects. Nevertheless, they played an important role in the first
days of computer art. Those days must be considered high days of creativity. Something
great was happening then, something took on shape. Not many had the guts to
clearly say this. It was happening at different places within a short time, and the activists
were not aware of each other. Yet, what they did, was of the same kind. They
surprised gallery owners who, of course, did not really like the art because, how
could they possibly make money with it? With the computer in the background, this
was mass production.
If the early pioneers themselves did not really understand the revolution they
were causing, they left art critics puzzling even more. “Is it or is it not art?” was
their typical shallow question, and: “Who (or what!) is the creator? The human, the
computer, or the drawing automaton?” The simplest of those first creations were
governed by polygons. Polygons became the signature of earliest algorithmic art.
This is why I tell their story.
In mathematics, a polygon is a sequence of points (in the simplest case, in the
plane). Polygons also exist in spaces of higher dimensions. As a sequence of points,
the polygon is a purely mental construct. In particular and against common belief,
you cannot see the polygon. As a polygon, it is invisible. It shares this fate with all
of geometry. This is so because the objects of geometry—points, lines, planes—are
pure. You describe them in formulae, and you prove theorems about them.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 65
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
I cannot avoid writing down how a point, a straight line, and a plane are given
explicitly. This must be done to provide a basis for the effort of an artist moving into
this field. So the point in three-dimensional space is an unrestricted triple of coordinates,
P = (x,y,z). The straight line is constructed from two points, say P1 and
P2, by use of one parameter, call it t. The values of t are real numbers, particularly
those between 0 and 1. The parameter acts like a coordinate along the straight line.
Thus, we can describe each individual point along the line by the formula
P(t) = P1 + t(P2 − P1). (3.1)
Finally, the points of a plane are determined from three given points by use of
two parameters:
P(u,v) = uP1 + vP2 + (1 − u − v)P3. (3.2)
We need two parameters because the plane is spreading out into two dimensions
whereas the straight line is confined to only one.
Bothering my readers with these formulae has the sole purpose that they should
become aware of the different kind of thinking required here. Exactly describing the
objects of hopefully ensuing creativity is only the start. It is parallel to the traditional
artist’s selection of basic materials. But algorithmic treatment must follow, if anything
is going to happen (we don’t do this here). The parameters u and v, I should
add, can be any real numbers. The three points are chosen arbitrarily, but then are
fixed (they must not be collinear).
As indicated above, all this is invisible. As humans, however, we want to see and,
therefore, we render polygons visibly. When we do so, we interpret the sequence of
points that make up the polygon, in an appropriate manner. The usual interpretation
is to associate with each point a location (in the plane or in space). Next, draw a
straight line from the first to the second point of the polygon, from there to the third
point, etc. A closed polygon, in particular, is one whose first and last points coincide.
To draw a straight line, of course, requires that you specify the colour and the
width of your drawing instrument, say a pencil. You may also want to vary the
strokeweight along the line, or use a pattern as you move on. In short, the geometry
and the graphics must be described explicitly and with utmost precision.
You have just learned your first and most important lesson: geometry is invisible,
graphics is visible. The entities of geometry are purely mental. They are related to
graphic elements. Only in them, they appear. Graphics is the human’s consolation
for geometry.
Let this be enough for a bit of formal and terminological background. We now
turn to the first years of algorithmic art.3 It is a well-established fact that between
3The art we are talking about, in the mid-1960s, was usually called computer art. This was certainly
an unfortunate choice. It used a machine, i.e. the instrument of the art, to define it. This had
not happened before in art history. Algorithmic art came much closer to essential features of the
aesthetic endeavour. It does so up to this day. Today, the generally accepted term is digital art. But
the digital principle of coding software is far less important than the algorithmic thinking in this
art, at least when we talk about creativity. The way of thinking is the revolutionary and creative
change. Algorithmic art is drawing and painting from far away.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
66 F. Nake
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
1962 and 1964 three mathematicians or engineers, who on their jobs had easy and
permanent access to computers, started to use those computers to generate simple
drawings by executing algorithms. As it happened, all three had written algorithms
to generate drawings and, without knowing of each other, decided to publicly exhibit
their drawings in 1965. Those three artists are (below, examples of their works will
be discussed):
• Georg Nees of Siemens AG, Erlangen, Germany, exhibited in the Aesthetic Seminar,
located in rooms of the Studiengalerie of Technische Hochschule Stuttgart,
Germany, from 5 to 19 February, 1965. Max Bense, chairing the institute, had
invited Nees. A small booklet was published as part of the famous rot series for
the occasion. It most likely became the first publication ever on visual computer
art (Nees and Bense 1965).4
• A. Michael Noll of Bell Telephone Laboratories, Murray Hill, NJ, USA showed
his works at Howard Wise Gallery in New York, NY, from 6 to 24 April, 1965
(together with random dot patterns for experiments on visual perception, by Bela
Julesz; the exhibits were mixed with those of a second exhibition).
• Frieder Nake from the University of Stuttgart, Germany, displayed his works at
Galerie Wendelin Niedlich in Stuttgart, from 5 to 26 November, 1965 (along with
Georg Nees’ graphics from the first show). Max Bense wrote an introductory
essay (but could not come to read it himself).5
As it happens, there may have been one or two forgotten shows of similar productions.6
But these three shows are usually cited as the start of digital art. The
public appearance and, thereby, the invitation of critique, is the decisive factor if
what you do is to be accepted as art. The artist’s creation is one thing, but only a
public reaction and critique can evaluate and judge it. The three shows, the authors,
and the year define the beginning of algorithmic art.
From the point of view of art history, it may be interesting to observe that conceptual
art and video art had their first manifestations around the same time. Op art had
existed for some while before concrete and constructive art became influential. The
happening—very different in approach—had its first spectacular events in the 1950s,
4The booklet, rot 19, contains the short essay, Projekte generativer Ästhetik, by Max Bense. I consider
it to be the manifesto of algorithmic art, although it was not expressly called so. It has been
translated into English and published several times. The term generative aesthetics was coined
here, directly referring to Chomsky’s generative grammar. The brochure contains reproductions of
some of Nees’ graphics, along with his explanations of the code.
5Bense’s introductory text, in German, was not published. It is now available on the compArt Digital
Art database at compart-bremen.de. Concerning the three locations of these 1965 exhibitions,
Howard Wise was a well-established New York gallery, dedicated to avant-garde art. Wendelin
Niedlich was a bookstore and gallery with a strong influence in the Southwest of Germany. The
Studiengalerie was an academic (not commercial) institution dedicated to experimental and concrete
art.
6Paul Brown recently (2009) discovered that Joan Shogren appears to have displayed computergenerated
drawings for the first time on 6 May 1963 at San Jose State University.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 67
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
Fig. 3.1 Georg Nees: 23-Ecke, 1965 (with permission of the artist)
and was continuing them. Pop art was, of course, popular. Serial, permutational, random
elements and methods were being explored by artists. Kinetic art and light art
were another two orientations of strong technological dependence. Max Bense had
chosen the title Programming the beautiful (Programmierung des Schönen) for the
third volume of his Aesthetica (Bense 1965), and Karl Gerstner had presented his
book Designing Programs (Programme entwerfen, Gerstner 1963), whose second
edition already contained a short section on randomness by computers.
But back to polygons! They appear in the works of the three above mentioned
scientists-turned-artists among their very first experiments (Figs. 3.1, 3.2 and 3.3).
We will now look at some of their commonalities and differences.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
68 F. Nake
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
Fig. 3.2 A. Michael Noll:
Gaussian-Quadratic, 1965
(with permission of the artist)
Assume you have at your disposal a technical device capable of generating drawings.
Whatever its mode of operation may be, it is a mechanism whose basic and
most remarkable operation creates a straight line-segment between two points. In
such a situation, you will be quite content using nothing but straight lines for your
aesthetic compositions. What else could you do? In a way, before giving up, you are
stuck with the straight line, even if you prefer beautifully swinging curved lines.
At least for a start you will try to use your machine’s capability to its very best
before you begin thinking about what other and more advanced shapes you may
be able to construct out of straight line-segments. Therefore, it was predictable (in
retrospect, at least) that Nees, Noll, and Nake would come up with polygonal shapes
of one or the other kind.
A first comment on creativity may be in order here. We see, in those artists’ activities,
the machinic limitations of their early works as well as their creative transcendence.
The use of the machine: creative. The first graphic generations: boring. The
use of short straight line-segments to draw bending curves: a challenge in creative
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 69
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
Fig. 3.3 Frieder Nake: Random Polygon, 1965
use of the machine. Turning to mathematics for the sake of art: creative, as well as
nothing particularly exciting. Throughout the centuries, many have done this. But
now the challenge had become to make a machine draw, whose sole purpose was
calculation. How to draw when your instrument is not made for drawing?
3.2.1 Georg Nees
Although “polygons” were Nees’, Noll’s, and Nake’s common first interest, their
particular designs varied considerably. In six lines of ordinary German text, Nees
describes what the machine is supposed to do (Nees and Bense 1965). An English
translation of his pseudo-code reads like this:
Start anywhere inside the figure’s given square format, and draw a polygon of 23 straight
line segments. Alternate between horizontal and vertical lines of random lengths. Horizontally
go either left or right (choose at random), vertically go up or down (also random
choice). To finish, connect start and end points by an oblique straight line.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
70 F. Nake
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
Clearly, before we reach the more involved repetitive design of Fig. 3.1, this
basic design must be inserted into an iterative structure of rows and columns. Once
a specific row and a specific column have been selected, the empty grid cell located
there will be filled by a new realisation of the microstructure just described. As we
see from the figure, the composition of this early generative drawing is an invisible
grid whose cells contain random 23-gons.
The random elements of Nees’ description of the polygon guarantee that, in all
likelihood, it will take thousands of years before a polygon will appear equal to, or
almost equal to, a previous one. The algorithm creates a rich and complex image, although
the underlying operational description appears as almost trivial. The oblique
line connecting the first and last points adds a lot to the specific aesthetic quality of
the image. It is an aberration from the rectilinear and aligned geometry of the main
part of the polygons. This aberration from a standard is of aesthetic value: surprise.
There are 19 × 14 = 266 elementary figures arranged into the grid structure.
Given the small size of the random shapes, we may, perhaps, not immediately perceive
polygons. Some observers may identify the variations on a theme as a design
study of a vaguely architectural kind.
The example demonstrates how a trivial composition can lead to a mildly interesting
visual appearance not void of aesthetic quality. I postpone the creativity issue
until we have studied the other two examples.
When some variable’s value is chosen “at random”, and this is happening by running
a computer program, the concept of randomness must be given an absolutely
precise meaning. Nothing on a computer is allowed to remain in a state of vagueness,
even if vagueness is the expressed goal. And even if the human observer of
the event does not see how he could possibly predict what will happen next, from
the computer’s position the next step must always be crystal clear. It must be computable,
or else the program does nothing.
In mathematics, a random variable is a variable that takes on its values only
according to a probability distribution. The reader no longer familiar with his or
her highschool mathematics may recall that a formula like y = x2 will generate the
result y = 16 if x = 4 is given. If randomness plays a role, such a statement could
only be made as a probability statement. This means the value of 16 may appear
as the result of the computation, but maybe it does not, and the result is, say, 17 or
15.7.
Usually, when in a programming language you have a function that, according to
its specification, yields random numbers, these numbers obey a so-called uniform
probability distribution. In plain terms, this says that all the possible events of an
experiment (like those of throwing dice) appear with the same probability.
But a random variable must not necessarily be uniformly distributed. Probability
distributions may be more complex functions than the uniform distribution. In early
algorithmic art, even of the random polygon variety, other distributions soon played
some role. They simulated (in a certainly naïve way) the artist’s intuition. (Does this
sound like too bold a statement?)
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 71
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
3.2.2 A. Michael Noll
A. Michael Noll’s “Gaussian-Quadratic” graphic makes use, in one direction (the
horizontal, viz. Fig. 3.2), of the Gaussian distribution. The coordinates of vertices in
the horizontal x-direction are chosen according to a Gaussian distribution, the most
important alternative to the uniform distribution. The co-ordinates of vertices in vertical
direction are calculated in a deterministic way (their values increase quadrati-
cally).
Whereas Nees’ design follows a definite, if simple, compositional rule, Noll’s is
really basic: one polygon whose points are determined according to two distributions.
It is not unfair to say that this is a simple visualisation of a simple mathematical
process.
3.2.3 Frieder Nake
The same is true of Nake’s polygon (Fig. 3.3). The algorithmic principle behind the
visual rendition is exactly the same as that of Fig. 3.2: repeatedly choose an x- and
a y-coordinate, applying distribution functions Fx and Fy, and draw a straight line
from the previous point to the new point (x,y); let then (x,y) take on the role of
the previous point for the next iteration.
In this formulation, Fx and Fy stand for functional parameters that must be provided
by the artist when his intention is to realise an image by executing the algorithm.7
Some experience, intuition, or creativity—whatever you prefer—flows into
this choice.
The visual appearance of Nake’s polygon may look more complex, a bit more
like a composition. The fact that it owes its look to the simple structure of one polygon,
does not show explicitly. At least, it seems to be difficult to visually follow
the one continuous line that constitutes the entire drawing. However, we can clearly
discover the solitary line, when we read the algorithm. The description of the simple
drawing contains more (or other) facts than we see. So the algorithmic structure
may disappear behind the visual appearance even in such a trivial case. Algorithmic
simplicity (happening at the subface of the image, its invisible side) may generate
visual complexity (visible surface of the image). If this is already happening
in such trivial situations, how much more should we expect a non-transparent relation
between simplicity (algorithmic) and complexity (visual) in cases of greater
algorithmic effort?8
7Only a few steps must be added to complete the algorithm: a first point must be chosen, the total
number of points for the polygon must be selected, the size of the drawing area is required, and the
drawing instrument must be defined (colour, stroke weight).
8The digital image, in my view, exists as a double. I call them the subface and the surface. They
always come together, you cannot have one without the other. The subface is the computer’s view,
and since the computer cannot see, it is invisible, but computable. The surface is the observer’s
view. It is visible to us.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
72 F. Nake
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
This first result occurred at the very beginning of computer art. It is, of course,
of no surprise to any graphic artist. He has experienced the same in his daily work:
with simple technical means he achieves complex aesthetic results. The rediscovery
of such a generative principle in the domain of algorithmic art is remarkable only
insofar as it holds.
However, concerning the issue of creativity, some observers of early algorithmic
experiments in the visual domain immediately started asking where the “generative
power” (call it “creativity”, if you like) was located. Was it in the human, in the
program, or even in the drawing mechanism? I have never understood the rationale
behind this question: human or machine—who or which one is the creator? But
there are those who love this question.
If you believe in the possibility of answering such a question, the answer depends
on how we first define “creative activity”. But such a hope usually causes us to
define terms in a way that the answer turns out to be what we want it to be. Not an
interesting discussion.
When Georg Nees had his first show in February 1965, a number of artists had
come to the opening from the Stuttgart Academy of Fine Art. Max Bense read his
text on projects of generative aesthetics, before Nees briefly talked about technical
matters of the design of his drawings and their implementation. As he finished, one
of the artists got up and asked: “Very fine and interesting, indeed. But here is my
question. You seem to be convinced that this is only the beginning of things to
come, and those things will be reaching way beyond what your machine is already
now capable of doing. So tell me: will you be able to raise your computer to the
point where it can simulate my personal way of painting?”
The question appeared a bit as if the artist wanted to give a final blow to the programmer.
Nees thought about his answer for a short moment. Then he said: “Sure,
I will be able to do this. Under one condition, however: you must first explicitly tell
me how you paint.” (The artists appeared as if they did not understand the subtlety
and grandeur, really: the dialectics of this answer. Without saying anything more,
they left the room under noisy protest.)
When Nietzsche, as one of the earliest authors, experienced the typewriter as a
writing device, he remarked that our tools participate in the writing of our ideas.9
I read this in two ways. First, in a literal sense. Using a pencil or a typewriter in
the process of making ideas explicit by formulating them in prose and putting this
in visible form on paper, obviously turns the pencil or typewriter in my hand into
a device without which my efforts would be in vain. This is the trivial view of the
tool’s involvement in the process of writing.
The non-trivial view is the observation that my thinking and attitude towards the
writing process and, therefore, the content of my writing is influenced by the tool
I’m using. My writing changes not only mechanically, but also mentally, depending
on my use of tools. It still remains my writing. The typewriter doesn’t write anything.
9Friedrich Kittler quotes Nietzsche thus: “Unser Schreibzeug arbeitet mit an unseren Gedanken.”
(Our writing tools participate in the writing of our thoughts.) (Kittler 1985), cf. Sundin (1980).
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 73
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
It is me who writes, even though I write differently when I use a pen than when I
use a keyboard.
The computer is not a tool, but a machine, and more precisely: an automaton.10
I can make such a claim only against a position concerning tools and machines and
their relation. Both, machines and tools, are instruments that we use in work. They
be-long to the means of any production. But in the world of the means of production,
tools and machines belong to different historic levels of development. Tools appear
early, and long before machines. After the machine has arrived, tools are still with
us, and some tools are hard to distinguish from machines. Still, to mix the two—as
is very popular in the computing field where everything is called a “tool”—amounts
to giving up history as an important category for scientific analysis. Here we see
how the ideological character of so many aspects of computing presents itself.
Nietzsche’s observation, that the tools of writing influence our thoughts, remains
true. Using the typewriter, he was no longer forced to form each and every letter’s
shape. His writing became typing: he moved from the continuous flow of the arm
and hand to the discrete hits of the fingers. We discover the digital fighting the
analog: for more precision and control, but also for standardisation. Similarly, I give
up control over spelling when I use properly equipped software (spell-checker). At
the same time, I gain the option of rapid changes of typography and page layout.
If creation is to generate something that was not there before, then it is me who
is creative. My creation may dwell on a trivial level. The more trivial, the easier it
may be to transfer some of my creative operations onto the computer. It makes a
difference to draw a line by hand from here to roughly there on a sheet of paper,
as compared to issuing the appropriate command sequence, which I know connects
points A and B. My thought must change. From “roughly here and there” to “precisely
these coordinates”.
My activity changes. From the immediate actor and generator of the line, I transform
myself into the mediating specifier of conditions a machine has to obey when
it generates the physical line. My part has become “drawing by brain” instead of
“drawing by hand”. I have removed myself from the immediacy of the material.
I have gained a higher level of semioticity.
My brain helps me to precisely describe how to draw a line between any two
points, whereas before I always drew just one line. It always was a single and particular
line: this line right here. Now it has become: this is how you do it, independent
of where you start, and where you end. You don’t embark on the adventure of
actually and physically drawing one and only one line. You anticipate the drawing
of any line.
I am the creative one, and I remain the creator. However, the stuff of my creation
has changed from material to semiotic, from particular to general, from single case
to all cases. As a consequence, my thinking changes. I use the computer to execute a
program. This is an enormous shift from the embodied action of moving the pencil.
Different skills are needed, different thinking is required and enforced. Those who
10Cf. Sundin (1980).
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
74 F. Nake
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
claim the computer has become creative (if they do exist) have views that appear
rather traditional. They do not see the dramatic change in artistic creation from
material to sign, from mechanics to deliberate semiotics.
What is so dramatic about this transformation? Signs do not exist in the world.
Other than things, signs require the presence of human beings to exist. Signs are
established as relations between other entities, be they physical or mental. In order
to appear, the sign must be perceived. In order to be perceivable, it must come in
physical form. That form, however, necessary as it is, is not the most important
correlate of the sign. Perceivable physical form is the necessary condition of the
sign; the full sign, however, must be constituted by a cognitive act.
Semiotics is the study of sign processes in all their multitudes and manifestations.
One basic question of semiotics is: how is communication possible? Semiotic
answers to this question are descriptive, not explanatory.
3.3 The Second Narration: On Three Artists
It has often been pointed out that computer art originates in the work of mathematicians
and engineers. Usually, this is uttered explicitly or implicitly with an undertone
on “only mathematicians and engineers”.
The observation is true. Mathematicians and engineers are the pioneers of algorithmic
art, but what is the significance of this observation? Is it important? What
is the relevance of the “only mathematicians” qualification? I have always felt that
this observation was irrelevant. It could only be relevant in a sense like: “early computer
art is boring; it is certainly not worth being called art; and no wonder it is so
boring—since it was not inspired by real artists, how could it be exciting”?
Frankly, I felt insulted a bit by the “only mathematicians” statement.11 It implies
a vicious circle. If art is only what artists generate, then how do you become an artist,
if you are not born an artist? The only way out of this dilemma is that everyone is,
in fact, born an artist (as not only Joseph Beuys has told us). But then the “only
mathematicians” statement wouldn’t make sense any more.
People generate objects and they design processes. They do not generate art. Art,
in my view, is a product of society—a judgement. Without appearing in public and
thus without being confronted with a critique of historic and systematic origin, a
work remains a work, for good or bad, but it cannot be said to have been included in
the broad historic stream of art. Complex processes take place after a person decides
to display his or her product in publicly accessible spaces. It is only in the public
domain that art can emerge (as a value judgement!). Individuals and institutions in
mutual interdependence are part of the processes that may merge to the judgement
that a work is assessed and accepted as a work of “art”—often enough, as we all
know, sparking even more controversy.
11This should read “mathematicians or engineers”, but I will stick to the shorter version.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 75
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
In the course of time, it often happens that an individual person establishes herself
or himself stably or almost irrevocably in the hall of art. Then she or he can
do whatever they want to do, and still get it accepted as “art”. But the principle
remains.12
The “only mathematician” statement is relevant only insofar as it is interpreted as
“unfortunately the pioneers were only mathematicians. Others did not have access
to the machines, or did not know how to program. Therefore we got the straight-line
quality of early works.”
However, if we accept that a work’s quality as a work of art is judged by society
anyhow, the perspective changes. Mathematician or bohemian does not matter
then. There cannot be serious doubt that what those pioneering mathematicians did
caused a revolution. They separated the generation of a work from its conception.
They did this in a technical way. They were interested in the operational, not only
mental separation. No wonder that conceptual art was inaugurated at around the
same time. The difference between conceptual and computational art may be seen
in the computable concepts that the computer people were creating.
However, when viewed from a greater distance, the difference between conceptual
artists and computational artists is not all that great. Both share the utmost
interest in the idea (as opposed to the material), and Sol LeWitt was most outspoken
on this. The early discourse of algorithmic art was also rich about the immaterial
character of software. Immaterial as software may be, it does not make sense without
being executed by a machine. A traditionally described concept does not have
such a surge to execution.13
The pioneers from mathematics showed the world that a new principle had arrived
in society: the algorithmic principle! No others could have done this, certainly
not artists. It had to be done by mathematicians, if it was to be done at all. The parlance
of “only mathematicians” points back to the speaker more than to the mathe-
matician.
Trivial to note is that creative work in art, design, or any other field, depends on
ideas on one hand, and skills on the other. At times it happens that someone has
a great idea but just no way to realise it. He or she depends on others to do that.
Pushing things a bit to the extreme, the mathematics pioneers of digital art may not
have had great ideas, but they knew how to realise them.
12Marcel Duchamp was the first to talk and write about this: “All in all, the creative act is not
performed by the artist alone; the spectator brings the work in contact with the external world by
deciphering and interpreting its inner qualification and thus adds his contribution to the creative act.
This becomes even more obvious when posterity gives a final verdict and sometimes rehabilitates
forgotten artists.” (Duchamp 1959). This position implies that a work may be considered a work of
art for some while, but disappear from this stage some time later, a process that has often happened
in history. It also implies that a person may be considered a great artist only after his or her death.
That has happened, too.
13It is a simplification to concentrate the argument on conceptual vs. algorithmic artists. There
have been other directions for artistic experiments, in particular during the 1960s. They needed a
lot of technical skill and constructive intelligence or creativity. Recall op art, kinetic art, and more.
Everything that humans eventually transfer to a machine has a number of precursors.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
76 F. Nake
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
On the other hand, artists may have had great ideas and lots of good taste and
style, but no way of putting that into existence. So who is to be blamed first? Obviously,
both had to acquire new and greater knowledge, skills, and feelings. They
had to learn from each other. Turning the argument around, we come up with “unfortunately,
some were only artists and therefore had no idea how to do it.” Doesn’t
this sound stupid? It sounds as stupid the other way around.
So let us take a look at what happened when artists wanted, and actually managed,
to get access to computers. As examples I have chosen Vera Molnar, Charles
Csuri, and Manfred Mohr. Many others could be added. My intent, however, is not
to give a complete account, a few cases are enough to make the point.
3.3.1 Vera Molnar
Vera Molnar was born in Hungary in 1924 and lived in Paris. She worked on concrete
and constructive art for many years. She tried to introduce randomness into her
graphic art. To her great dismay, however, she realised that it is hard for a human to
avoid repetition, clusters, trends, patterns. “Real” randomness does not seem to be
a human’s greatest capability.
So Vera Molnar decided that she needed a machine to do parts of her job. The
machine would not be hampered by the human subjectivity that seems to get in
the way of a human trying to do something randomly. The kind of machine she
needed was a computer that, of course, she had no access to. Vera Molnar felt that
systematic as well as hazardous ways of expressing and researching were needed
for her often serial and combinatorial art. Since she did not have the machine to
help her to do this, she had to build one herself. She did it mentally: “I imagined I
had a computer” (Herzogenrath and Nierhoff 2006, p. 14). Her machine imaginaire
consisted of exactly formulated rules of behaviour. Molnar simulated the machine
by strictly doing what she had told the imaginary machine to do.
In 1968, Vera Molnar finally gained access to a computer at the Research Centre
of the computer manufacturer, Bull. She learned programming in Fortran and Basic,
but also had people to help her. She did not intend to become an independent programmer.
Her interests were different. For her, the slogan of the computer as a tool
appears to be justified best. She allowed herself to change the algorithmic works by
hand. She made the computer do what she did not want to do herself, or what she
thought the machine was doing more precisely.14
Figure 3.4 (left)15 shows one of her early computer works. She had previously
used repertoires of short strokes in vertical, horizontal, or oblique directions, sim-
14The catalogue (Herzogenrath and Nierhoff 2006) contains a list of the hardware Vera Molnar has
used since 1968. It also presents a thorough analysis of her artistic development. The catalogue
appeared when Molnar became the first recipient of the d.velop digital art award. A great source
for Molnar’s earlier work is Hollinger (1999).
15This figure consists of two parts: a very early work, and a much later one by the same artist. The
latter one is given without any comment to show an aspect of the artist’s development.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 77
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
Fig. 3.4 Vera Molnar. Left: Interruptions, 1968/69. Right: 25 Squares, 1991 (with permission of
the artist)
ilar in style to what many of the concrete artists had also done. The switchover
to the computer gave her the opportunity to do more systematic research. (“Visual
research” was a term of the time. The avantgarde loved it as a wonderful shield
against the permanent question of “art”. Josef Albers and others from the Bauhaus
were early users of the word.)
The Interruptions of Fig. 3.4 happen in the open spaces of a square area that is
densely covered by oblique strokes. They build a complex pattern, a texture whose
algorithmic generation, simple as it must be, is not easy to identify. The open areas
appear as surprise. The great experiment experienced by pioneers of the mid-1960s
shows in Molnar’s piece: what will happen visually if I force the computer to obey a
simple set of rules that I invent? How much complexity can I generate out of almost
trivial descriptions?
3.3.2 Charles Csuri
Our second artist who took to the computer is Charles Csuri. He is a counter example
to the “only mathematicians” predicament. Among the few professional artists
who became early computer users, Csuri was probably the first. He had come to
Ohio State University in Columbus from the New York art scene. His entry into the
computer art world was marked by a series of exceptional pieces, among them Sine
Curve Man (Fig. 3.5, left), Random War, and the short animated film Hummingbird
(for more on Csuri and his art, see Glowski 2006).
Sine Curve Man won him the first prize of the Computer Art Contest in 1967.
Ed Berkeley’s magazine, Computers and Automation (later renamed to Computers
and People), had started this yearly contest. It was won in 1965 by A. Michael Noll,
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
78 F. Nake
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
Fig. 3.5 Charles Csuri. Left: Sine Curve Man, 1967. Right: yuck 4x3, 1991 (with permission of
the artist)
1966 by Frieder Nake, and then by Csuri, an educated artist for the first time. This
award, by the way, never gained high esteem. It took many more years, until 1987,
when the now extremely prestigious Prix Ars Electronica was awarded for the first
time.
For his first programming tasks, Csuri was assisted by programmer James Shaffer.
Similar to Vera Molnar, we see that the skill of programming may at the beginning
constitute a hurdle that is not trivial to master. If time plays a role, an artist
willing to use the computer, but still unable to do it all by himself, has almost no
choice but to rely on friendly cooperation. Such cooperation may create friction with
all its negative effects. As long as the technical task itself does not require cooperation,
it is better to acquire the new technical skill. After all, there is no art without
skillful work, and a steadily improved command of technical skills is a necessary
condition for the artist. Why should this be different when the skill is not the immediate
transformation of a corporeal material by hand, but instead the description
only of relations and conditions, of options and choices of signs?
Csuri’s career went up steeply. Not only did he become the head of an academic
institute but even an entrepreneur. At the time of a first rush for large and leading
places in computer animation, when this required supercomputers of the highest
technological class and huge amounts of money, he headed the commercial Cranston
Csuri Productions company as well as the academic Advanced Computing Center
for the Arts and Design, both at Columbus, Ohio. In the year 2006, Csuri was honoured
by a great retrospective show at the ACM SIGGRAPH yearly conference.
Sine Curve Man is an innovation to computer art of the first years in two respects:
its subject matter is figurative, and it uses deterministic mathematical techniques
rather than probabilistic. There is a definite artistic touch to the visual appearance
of the graphic (Fig. 3.5), quite different from the usual series of precise geometric
curves that many believe computer art is (or was) about.
The attraction of Sine Curve Man has roots in the graphic distortions of the (old?)
man’s face. Standard mathematics can be used for the construction. A lay person
may, however, not be familiar with such methods. Along the curves of an original
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 79
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
drawing, a series of points are marked. The curves may, perhaps, have been extracted
from a photograph. The points become the fixed points of interpolations by sums of
sine functions. This calculation, purely mathematical as it is, and without any intuitive
additions triggered by the immediate impression of a seemingly half-finished
drawing, is an exceptional case of the new element in digital art.
This element is the dialectics of aesthetics and algorithmics. Sine Curve Man
may cause in an observer the impression that something technical is going on. But
this is probably not the most important aspect. More interesting is the visual (i.e.
aesthetic) sensation. The distortions this man has suffered are what attracts us. We
are almost forced to explore this face, perhaps because we want to read the curves
as such. But they do not allow us to do this. Therefore, our attention cannot rest with
the mathematics. Dialectics happens, as well as semioses (sign processes): jumping
back and forth between semantics and syntactics.
3.3.3 Manfred Mohr
Manfred Mohr is a decade younger than the first two artists. They belong to the first
who were accepted by the world of art despite their use of computers. Do they owe
anything to computers? Hard to say. An art historian or critic will certainly react
differently if he doesn’t see an easel in the artist’s studio, but a computer instead.
The artist doesn’t owe much to a computer. He has decided to use it, whatever
the reason may have been. If to anything, he owes to the programs he is using or
has written himself. With those programs, he calls upon work formerly spent that he
now is about to set in action again. The program appears as canned labour ready to
be resuscitated.
The relation between artist and computer is, at times, romanticised as if it were
similar to the close relation between the graphic artist and her printer (a human
being). The printer takes pride in getting the best quality out of the artist’s design.
The printing job takes on artistic quality itself. The computer, to the contrary, is
only executing a computable function. It should be clear, that the two cases are as
different as they could ever be.
If we characterise Vera Molnar, in one word, as the grand old lady of algorithmic
art, and Charles Csuri as the great entrepreneur and mover, Manfred Mohr would
appear as the strictest and strongest creator of a style in algorithmic art. The story
says that his young and exciting years of searching for his place in art history were
filled with jamming the saxophone, hanging out in Spain and France, and with hard
edge constructivist paintings. Precision and rationality became and remained his
values. They find a correspondence and a balancing force in the absolute individual
freedom of jazz. Like many of the avant-garde artists in continental Europe during
the 1960s, he was influenced by Max Bense’s theory and writing on aesthetics, and
when he read in a German news magazine (Anon 1965) that computers had appeared
in fine art, he knew where he had to turn to.
K.R.H. Sonderborg and the art of Informel, Pierre Barbaud and electronic music,
Max Bense and his theory of the aesthetic object constitute a triad of influences from
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
80 F. Nake
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
Fig. 3.6 Manfred Mohr. Left: P-18 (Random Walk), 1969. Right: P-707-e1 (space.color),
1999–2001 (with permission of the artist)
which Mohr’s fascinating generative art emerged. From his very first programmed
works in 1969 to current days, he has never betrayed his striving for the greatest
transparency of his works. Never did he leave any detail of his creations open to
hand-waving or to dark murmurs. He discovered the algorithmic description of the
generative process as the new creation. The simplest elements can become the material
for the most complex visual events.
After about four years of algorithmic experiments with various forms and relations,
Manfred Mohr, in 1973, decided to use the cube as the source of external
inspiration. He has continued exploring it ever since. There are probably only a few
living persons who have celebrated and used the cube more than him (for further
information see Keiner et al. 1994, Herzogenrath et al. 2007).
Figure 3.6 shows one event in the six-dimensional hypercube (right), and one of
the earliest generative graphics of Mohr’s career (left).
When we see a work by Mohr, we immediately become aware of the extraordinary
aesthetic quality of his work. His decisions are always strong and secure. The
random polygon of Fig. 3.6 is superior to most, if not all, of the others one could
see in the five years before. The events of the heavier white lines add an enormous
visual quality to the drawing, achieved in such strength here for the first time.
The decision, in 1973, to explore the three-dimensional cube as a source for
aesthetic objects and processes, put Manfred Mohr in a direct line with all those
artists who, at least for some part of their artistic career, have explored one and the
same topic over and over again. It should be emphasised, however, that his interest
in the cube and the hypercube16 does not signify any pedagogical motif. He does not
intend to explain anything about spaces of higher dimensions, nor does he visualise
16The hypercube is analogous to a three-dimensional cube in four or more dimensions. It is recursively
defined as an intricate structure of cubes.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 81
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
cubes in six or eleven dimensions. He takes those mental creatures as the rational
starting points for his visual creation. The hypercube is only instrumental in Mohr’s
creative work; it is not the subject matter.
The cube in four or more dimensions is a purely mental product. We can clearly
think the hypercube. But we cannot visualise it. We may take the hypercube as
the source of visual aesthetic events (and Mohr does it). But we cannot show it in
a literal sense of the word. Manfred Mohr’s mental hikes in high dimensions are
his inspiration for algorithmic concrete images. For these creations, he needs the
computer. He needs it even more when he allows for animation.
Manfred Mohr’s work stands out so dramatically because it cannot be done without
algorithms. It is the most radical realisation of Paul Klee’s announcement: we
don’t show the visible, we make visible. The image is a visible signal. What it shows
is itself. It has a source elsewhere. But the source is not shown. It is the only reason
for something visible.
Creativity? Yes, of course, piles of. Supported by computer? Yes, of course, in
the trivial sense that this medium is needed for the activity of realising something
the artist is thinking of. In Manfred Mohr’s work (and that of a few others whose
number is increasing) generative art has actually arrived. The actuality of his work
is its virtuality.
3.4 The Third Narration: On Two Programs
Computer programs are, first of all, texts. The text describes a complex activity. The
activity is usually of human origin. It has before existed as an activity carried out
by humans in many different forms. When it becomes the source of an algorithmic
description, it may gradually disappear as a human activity, until in the end, the
computer’s (or rather the program’s) action appears as the first and more important
than the human activities that may still be needed to keep the computer running:
human-supported algorithmic work.
The activity described by a computer program as a text may be almost trivial, or
it may be extremely complex. It may be as trivial as an approximate calculation of
the sine function for a given argument. Or it may be as complex as calculating the
weather forecast for the area of France by taking into account all available atmospheric
measurements collected around the world.
The art of writing computer programs has become a skill of utmost creativity, intuition,
constructive precision, and secrets of the trade. Donald Knuth’s marvellous
series of books, The Art of Computer Programming, is the best proof of this (Knuth
1968). These books are one of the greatest attempts to give an in-depth survey of
the entire field of computing. It is almost impossible to completely grasp this field
in totality, or even to finish writing the series of books. Knuth is attempting to do
just this.
Computer programs have been characterised metaphorically as tools, as media,
or as automata. How can a program be an automaton if it is, as I have claimed,
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
82 F. Nake
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
a text? The answer is in the observation that the computer is a semiotic machine
(Nadin 2011, Nöth 2002, Nake 2009).
The computer is seen by these authors as a semiotic machine, because the stuff
it processes is of a semiotic nature. When the computer is running, i.e. when it is
working as a machine, it is executing a program. It is doing this under the control
of an operating system. The operating system is itself a program. The program,
that the computer is executing, takes data and transforms it into new data. All these
creatures—the operating system, the active program, and data—are themselves of
semiotic nature. This chapter is not the place to go deeper into the semiotic nature
of all entities on a computer.17 So let us proceed from this basic assumption.
The assumption becomes obvious when we take a look at a program as a text.
Leaving aside all detail, programming starts from a more or less precise specification
of what a program should be doing. Then there is the effort of a group of
programmers developing the program. Their effort materialises in a rich mixture of
activities. Among these, the writing of code is central. All other kinds of activities
eventually collapse into the writing of code.
The finished program, which is nothing but the code for the requested function,
appears as a text. During his process of writing, the programmer must read the text
over and over again. And here is the realisation: the computer is also reading the
text! The kind of text that we call “computer program” constitutes a totally new
kind of poetry. The poetics of this poetry reside in the fact that it is written for two
different readers: one of them human, the other machine.
Their fantastic semiotic capabilities single out humans from the animal kingdom.
Likewise, the computer is a special machine because of its fantastic semiotic
capabilities. Semiotic animal and semiotic machine meet in reading the text that is
usually called a program.
Now, reading is essentially interpreting. The human writer of the program materialises
in it the specification of some complex activity. During the process of his
writing, he is constantly re-reading his product as he has so far written it. He is
convinced of the program’s “correctness”. It is correct as long as it does what it is
supposed to do. However, how may a text be actively doing anything?
The text can do something only if the computer is also reading it. The reading,
and therefore interpreting, of the program by the computer effectively transforms the
text into a machine. The computer, when reading the program text (and therefore:
interpreting it), cannot but execute it. Without any choice, reading, interpreting, and
executing the text are one and the same for the computer. The program as a text
is interesting for the human only insofar as the computer is brought to execute it.
During execution, the program reveals its double character as text-and-machine,
both at the same time. So programs are executable texts. They are texts as machine,
and machine as text.
After this general but also concrete remark about what is new in postmodern
times, we take a look at two specific and ambitious, albeit very different programs.
17A book is in preparation that takes a fundamental approach to this topic: P.B. Andersen &
F. Nake, Computers and signs. Prolegomena to a semiotic foundation of computing.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 83
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
We don’t look at their actual code because this is not necessary for our discussion
of creativity in early computer art. Harold Cohen’s famous AARON started its astonishing
career in 1973, and continued to be developed for decades. Frieder Nake’s
Generative Aesthetics I was written, completed, then discarded in the course of one
year, 1968/69.
3.4.1 Harold Cohen: AARON
AARON is a rule-based system, an enormous expert system, one of the very few
expert systems that ever made it to their productive phase (McCorduck 1990). In the
end it consisted of so many rules that its sole creator, Cohen, was no longer sure if
he was still capable of understanding well enough their mutual dependencies.
Everything on a computer must be rule-based. A rule is a descriptive element
of the structure: if C then A, where C is a condition (in the logical sense of
“condition”), and A is an action. In the world of computing, a formal definition
must state precisely what is accepted as a C, and what is accepted as an A. In
colloquial terms, an example could be: if (figure ahead) then (turn
left or right). Of course, the notions of “figure”, “ahead”, “turn”, “left”,
“right” must also be described in computable terms, before this can make any sense
to a computer.
A rule-based system is a collection of interacting rules. Each rule is constructed
as a pair of a condition and an action. The condition must be a description of an event
depending on the state (value) of some variables. It must evaluate to one of the truthvalues
true or false. If its value is true, the action is executed. This requires
that its description is also given in computable form. The set of rules making up a
rule-based system may be structured into groups. There must be an order according
to which rules are tested for applicability. One strategy is to apply the first applicable
rule in a given sequence of rules. Another one determines all applicable rules and
selects one of them.
Cohen’s AARON worked for many years during which it produced a large collection
of drawings. They were first in black and white. Later, Cohen coloured them
by hand according to his own taste or to codes also determined by AARON. The
last stage of AARON relied on a new painting machine. It was constructed such that
it could mimic certain painterly ways of applying paint to paper.
During more than three decades, AARON’s command of subjects developed
from collections of abstract shapes to evocations in the observer of rocks, birds, and
plants, and to figures more and more reminiscent of human beings. They gave the
impression of a spatial arrangement, although Cohen never really entered into three
dimensions. A layered execution of figures was sufficient to generate a low-level of
spatial impression.
Around the year 2005, Cohen became somewhat disillusioned with the figural
subjects he had gradually programmed AARON to better and better create. When he
started using computers and writing programs in the early 1970s, he was fascinated
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
84 F. Nake
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
Fig. 3.7 Harold Cohen. Left: Early drawing by AARON, with the artist. Right: Drawing by
AARON, 1992 (with permission of the artist)
by the problem of representation. His question then was: just how much, or little,
does it take before a human observer securely recognises a set of lines and colours as
a figure or pattern of something? How could a painting paint itself? (Cohen 2007).
But Harold Cohen has now stopped following this path any further. He achieved
more than anyone else in the world in terms of creating autonomous rule-based art
systems. He did not give up this general goal. He decided to return to pure form and
colour as the subject matter of his autonomous rule-based system.
For a computer scientist, there is no deep difference between an algorithm and
a rule-based system. As Cohen (2007) writes, it took him a while to understand
this. The difference is one of approach, not of the results. Different approaches may
still possess the same expressive power. As Cohen is now approaching colour again
in an explicitly algorithmic manner, he has shifted his view closer to the computer
scientist’s but without negating his deep insight into the qualities of colour as an
artist.
This is marvellous. After a long and exciting journey, it sheds light on the alleged
difference between two views of the world. In one person’s great work, in his
immediate activity, experience, and knowledge, the gap between the “two cultures”
of C.P. Snow fades. It fades in the medium of the creative activity of one person,
not in the complex management of interdisciplinary groups and institutes. The book
must still be written that analyses the Cohen decades of algorithmic art from the
perspective of art history.
Cohen’s journey stands out as a never again to be achieved adventure. He has
always been the lonely adventurer. His position is unique and singular. Artificial
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 85
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
Intelligence people have liked him. His experience and knowledge of rule-based
systems must be among the most advanced in the world. But he was brave enough
to see that in art history he had reached a dead-end. Observers have speculated
about when would AARON not only be Cohen’s favourite artist, but also its own
and best critic. Art cannot be art without critique. As exciting as AARON’s works
may be, they were slowly losing their aesthetic appeal, and were approaching the
only evaluation: oh, would you believe, this was done by computer? The dead-end.
Harold Cohen himself sees the situation with a bit more skepticism. He writes:
It would be nice if AARON could tell me which of them [its products] it thinks I should
print, but it can’t. It would be nice if it could figure out the implications of what it does so
well and so reliably, and move on to new definitions, new art. But it can’t. Do those things
indicate that AARON has reached an absolute limit on what computers can do? I doubt it.
They are things on my can’t-do-that list. . . (Cohen 2007).
The can’t-do-that list contains statements about what the computer can and what
it cannot do. During his life, Cohen has experienced how items had to be removed
from the list. Every activity that is computable must be taken from the list. There are
activities that are not computable. However, the statement that something cannot be
done by computer, i.e. is not computable, urges creative people to change the noncomputable
activity into a computable one. Whenever this is achieved after great
hardship, we don’t usually realise that a new activity, a computable one, has been
created with the goal in mind to replace the old and non-computable.
There was a time, when Cohen was said to be on his way to becoming the first
artist of whom there would still be new works in shows after his death. He himself
had said so, jokingly with a glass of cognac in hand. He had gone so far that such
a thought was no longer fascinating. The Cohen manifesto of algorithmic art has
reached its prediction.
But think about the controversial prediction once more. If true, would it not be
proof of the computer’s independent creativity? Clearly, Cohen wrote AARON, the
program, the text, the machine, the text-became-machine. This was his, Cohen’s
creative work. But AARON was independent enough to then get rid of Cohen, and
create art all by itself. How about this?
In a trivial sense, AARON is creative, but this creativity is a pseudo-creativity. It
is confined to the rules and their certainly wide spectrum of possibilities. AARON
will forever remain a technical system. Even if that system contained some metarules
capable of changing other rules, and meta-meta-rules altering the meta-rules
on the lower level, there would always be an explicit end. AARON would not be
capable of leaving its own confines. It cannot cross borders.
Cohen’s creativity, in comparison, stands out differently. Humans can always
cross borders. A revolution has happened in the art world when the mathematicians
demonstrated to the artists that the individual work was no longer the centre of
aesthetic interest. This centre had shifted to descriptions of processes. The individual
work had given way to the class of works. Infinite sets had become interesting, the
individual work was reduced to a by-product of the class. It has now become an
instance only, an index of the class it belongs to.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
86 F. Nake
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
No doubt, we need the instance. We want to literally see something of the class.
Therefore, we keep an interest in the individual work. We cannot see the entire
class. It has become the most interesting, and it has become invisible. It can only be
thought.
I am often confronted with an argument of the following kind. A program is
not embedded into anything like a social and critical system, and clearly, without a
critical component, it cannot leave borders behind. So wait, the argument says, until
programs are embedded the proper way.
But computers and programs don’t even have bodies. How then should they be
able to be embedded in such critical and social systems? Purpose and interest are
just not their thing. Don’t you, my dear friends, see the blatant difference between
yourself and your program, between you and the machine?
Joseph Weizenbaum dedicated much of his life to convincing others of this fundamental
difference. It seems to be very tough for some of us to accept that we are
not like machines and, therefore, they are not like us.
3.4.2 Frieder Nake: Generative Aesthetics I
A class of objects can never itself, as a class, appear physically. In other words, it
cannot be perceived sensually. It is a mental construct: the description of processes
and objects. The work of art has moved from the world of corporeality to the world
of options and possibilities. Reality now exists in two modes, as actuality and virtu-
ality.
AARON’s generative approach is activity-oriented. The program controls a
drawing or painting tool whose movements generate, on paper or canvas, visible
traces for us to see. The program Generative Aesthetics I, however, is algorithmoriented.
It starts from a set of data, and tries to construct an image satisfying conditions
that are described in the data.
You may find details of the program in Nake (1974, pp. 262–277). The goal of the
program was derived from the theory of information aesthetics. This theory starts
by considering a visual artefact as a sign. The sign is really a supersign because it is
usually realised as a structure of signs.
The theory assumes that there is a repertoire of elementary or primitive signs.
Call those primitive signs: s1,s2,...,sr . They must be perceivable as individual
units. Therefore, they can be counted, and relative frequencies of their occurrence
can be established. Call those frequencies, f1,f2,...,fr .
In information aesthetics, a schema of the signs with their associated relative
frequencies is called a sign schema. It is a purely statistical description of a class
of images. All those images belong to the class that use the same signs (think of
colours) with the same frequencies.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 87
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
In Shannon’s information theory, the statistical measure of information in a message
is defined as
H = −
r
i=1
pi logpi. (3.3)
The assumption for the derivation of this formula in Shannon and Weaver (1963)
is that all the pi are probabilities. They determine the statistical properties of a
source sending out messages that are constructed according to the probabilities of
the source.
This explanation may not mean much to the reader. For one, information theory
is no longer popular outside of certain technical contexts. Moreover, it was overestimated
in the days when the world was hoping for a great unifying theory. The
measure H gives an indication of what we learn when one specific event (out of a
set of possible events) has occurred, and we know what the other possible events
could be.
Take as an example the throwing of dice in a typical board game. As we know,
there are six possible events, which we can identify by the numbers 1, 2, 3, 4, 5,
and 6. Each one of the six events occurs with the same probability, i.e. 1/6. Using
Shannon’s formula for the information content of the source “dice”, we get
H = −log(1/6) = −(log1 − log6) = log6 ≈ 2.6 (3.4)
(the logarithm must be taken to the base of 2). The result is measured in bits and
must be interpreted thus: when one of the possible results of the throw has appeared,
we gain between two and three bits of information. This, in turn, says that between
two and three decisions of a “yes or no” nature have been taken. The Shannon
measure of information is a measure of the uncertainty that has disappeared when
and because the event has occurred.
Information aesthetics, founded by Max Bense and Abraham A. Moles (Bense
1965, Moles 1968) and further developed in more detail by others (Gunzenhäuser
1962, Frank 1964), boldly and erroneously ignored the difference between frequency
and probability. To repeat, probabilities of a sign schema characterise an
ideal source. Frequencies, however, are results of empirical measurement of several,
but only finitely many messages or events (images in our case). As such, frequencies
are only estimates for probabilities.
Information aesthetics wanted to get away from subjective value judgement. Information
aesthetic criteria were to be objective. Aspects of the observer were excluded,
at least in Max Bense’s approach. Empirical studies from the 1960s and later
were, however, not about aesthetic sources, but about individual pieces. In doing so,
the difference of theory and practice, of infinite class and individual instance, of
probability and frequency, had to be neglected by replacing theoretical probability
by observed frequency, thus pi = fi. This opened up the possibility to measure the
object without any observer being present. However, the step also gave up aesthetics
as the theory of sensual perception.
Now, the program Generative Aesthetics I accepted as input a set of constraints of
the following kind. For each sign (think of colour), a measure of surprise and a measure
of conspicuity (defined by Frank in 1964) could be constraint to an interval of
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
88 F. Nake
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
feasible values. Such requirements defined a set of up to 2r constraints. In addition,
the aesthetic measure that Gunzenhäuser had defined as an information-theoretic
analogue to Birkhoff’s famous but questionable measure of “order in complexity”
(Birkhoff 1931) could be required to take on a maximum or minimum value, relative
to the constraints mentioned before. Requesting a maximum to be the goal of
construction put trust on the formal definition of aesthetic measure actually yielding
a good or even beautiful solution. Requesting a minimum, to the contrary, did not
really trust the formalism.
With such a statement of the problem, we are right into mathematics. The problem
turns out to be a non-linear optimisation problem. If a solution is possible, it had
to be a discrete probability distribution. This distribution represents all images satisfying
the constraints. it was called “the statistical pre-selector,” since it was based
only on a statistical view of the image. In a second step, a topological pre-selector
took the sign schema of the previous step and created the image as a hierarchical
structure of colour distribution, according to the probabilities determined before.
The type of structure used for this construction of the image was, in computer
science, later called a quadtree. A quadtree divides an image into four quadrants of
equal size. The generative algorithm distributes the probabilities of the entire image
into the four smaller quadrants such that the sum total remains the same. With each
quadrant, the procedure is repeated recursively, until a quadrant is covered by one
colour only, or its size has reached a minimal length.
Generative Aesthetics I thus bravely started from specifying quantitative criteria
that an image was to satisfy. Once the discrete probability distribution was determined
as a solution to the set of criteria, an interesting process of many degrees
of freedom started to distribute the probabilities into smaller and smaller local areas
of the image but such that the global condition was always satisfied. Aesthetics
happened generatively and objectively, by running an automaton.
The program was realised in the programming language PL/I with some support
from Fortran routines. Its output was trivial but fast. I was working on this project
in Toronto in 1968/69. Since no colour plotter was available, I used the line printer
as output device. The program’s output was a list of measures from information
aesthetics plus a coded printout of the generated image. I used printer symbols to
encode the colours that were to be used for the image.
This generative process was very fast, which allowed me to run a whole series
of experiments. These experiments may constitute the only ones ever carried out
in the spirit of generative aesthetics based on the Stuttgart school of information
aesthetics. The program was intended to become the base for empirical research
into generative aesthetics. Regrettably, this was not realised.
With the help of a group of young artists, I realised by hand only two of the
printouts. From a printer’s shop we got a set of small pieces of coloured cardboard.
They were glued to a panel of size 128 × 128 cm. One of those panels has been
lost (Fig. 3.8). The other one is in the collection Etzold at Museum Abteiberg in
Mönchengladbach, Germany.
Besides the experience of solving a non-trivial problem in information aesthetics
by a program that required heuristics to work, I did this project more like a scientist
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 89
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
Fig. 3.8 Frieder Nake: Generative Aesthetics I, experiment 4a.1, 1969
than an artist. An artist would have organised, well in advance, a production site
to transform the large set of the generated raster images into a collection of works.
This collection would become the stock of an exhibition at an attractive gallery.
A catalogue would have been prepared with the images, along with theoretical and
biographical essays. Such an effort to propagate the most advanced and radically
rational generative aesthetics would have been worthwhile.
Instead, I think I am justified in concluding that this kind of formally defined
generative aesthetics did not work. After all, my experiments with Generative Aesthetics
I seemed to constitute an empirical proof of this.
Was I premature in drawing the conclusion? It was the time of Cybernetic
Serendipity in London, Tendencies 4, and later Tendencies 5 in Zagreb. In Europe
one could feel some low level, but increasing attention being paid to computer art.
A special show was in preparation for the 35th Biennale in Venice, bringing together
Russian constructivists, Swiss concrete artists, international computer artists,
and kids playing. Wasn’t this an indication of computer art being recognised and
accepted as art. Premature resignation? Creativity not recognised?
I am not so sure any more. As a testbed for series of controlled experiments
on the information-aesthetic measures suggested by other researchers, Generative
Aesthetics I may, after all, have possessed a potential that was not really fathomed.
The number of experiments was too small. They were not designed systematically.
Results were not analysed carefully enough. And other researchers had not been
invited to use the testbed and run their own, most likely very different, experiments.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
90 F. Nake
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
It may well be the case that the project should be taken up again, now under more
favourable conditions, and different awareness for generative design.
3.5 The Fourth and Last Narration: On Creativity
This chapter finds its origins in a Dagstuhl Seminar in the summer of 2009. Schloss
Dagstuhl is a beautiful location hidden in the Southwest of Germany, in the province
of Saarland. Saarland is one of the European areas where over centuries people from
different nations have mixed. After World War II, Saarland belonged to France for
some time until a public vote was taken (in 1955) about where people preferred
to live, in West Germany or France. Was their majority decision in favour of the
German side an act of collective creativity?
Mathematicians in Germany and beyond have had a wonderful institution ever
since 1944, the Mathematical Research Institute of Oberwolfach. It is located at
Oberwolfach in the Black Forest. Mathematicians known internationally for their
interest in a specialised field, meet there to pursue their work. They come in international
groups, with an open agenda leaving lots of time for spontaneous arrangements
of discussion, group work, and presentations.
The German Gesellschaft für Informatik, after having established itself as a powerful,
active, and growing scientific association in the field of computing, became
envious of the mathematicians and decided that they also wanted to have such a
well-kept, challenging and inviting site for scientific meetings of high quality. Soon
enough, they succeeded. Was this creativity or organisation?
So Dagstuhl became a place for scientists and others, from computer science and
neighbouring disciplines, to gather in a beautiful environment and work on issues
of a specialised nature. They are supposed to come up with findings that should
advance theory and practice of information technology in the broadest sense.
A week at a Dagstuhl seminar is a great chance to engage in something that we
usually find no opportunity to do. The topic at this particular occasion was computational
creativity—a topic of growing, if only vague interest these days.
Inspired by some of the debates at the seminar, I have tried in this chapter, to
recall a few aspects from the early history of algorithmic art as a case from the
fringes of computing that we would usually consider a case for creativity. We usually
assume that for art to emerge, creativity must happen. So if we see any reason to do
research into the relation between creativity and computers, a study of computer art
seems to be a promising case.
People are, of course, curious to learn about human creativity in general. A special
interest in the impact of computing on creativity must have its roots in the huge
machine. As already indicated, I see the computer as a semiotic machine. The subject
matter of computational processes must always already belong to the field of
semiotics. The subject matter computers work on is of a relational character more
than it is “thing-like”.
This important characteristic of all computing processes exactly establishes a
parallel between computable processes and aesthetic processes. But to the extent
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 91
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
that computable processes are carried out by machinery, those processes cannot really
reach the pragmatic level of semiosis. Pragmatics is central to purpose. Purpose
is what guides humans in their activities. The category of purpose is strongly connected
to interest.
I don’t think it could be proved—in a rigorous mathematical meaning of the
word “prove”—that machines do not (and can never) possess any form of interest
and, therefore, cannot follow a purpose. On the other hand, however, I cannot see
any common ground between the survival instinct governing us as human beings,
and the endless repetition of always the same, if complex, operations the machine is
so great and unique at. There is just nothing in the world that indicates the slightest
trace of an interest on behalf of the machine. Even without such proof, I do not see
any reason or situation where I would use a machine, and this machine developed
anything I would be prepared to accept as “interest” and, in consequence, a purposeful
activity.
What above I have called an interpretation by the machine is, of course, an interpretation
only in a purely formal sense of the word. Clearly, the agent of such
interpretation is a machine. As a machine, it is constructed in such a way that it has
no freedom of interpretation. The machine’s interpretation is, in fact, of the character
of a determination: it must determine the meaning of a statement in an operational
way. When it does so, it must follow strict procedures hard-wired into it (even if it
is a program called a compiler that carries out the process of determination). This
does not allow a comparison to human interpretation.
3.6 Conclusion
The conclusion of this chapter is utterly simple. Like any other tool, material, or
media, computer equipment may play important roles in creative processes. A human’s
creativity can be enhanced, triggered, or encouraged in many ways. But there
is nothing really exciting about such a fact other than that it is rather new, it is extremely
exciting, it opens up huge options, and it may trigger super-surprise.
In the year 1747, Julien Offray de La Mettrie published in Leiden, the Netherlands,
a short philosophical treatise under the title L’Homme Machine (The Human
Machine).18 This is about forty years before the French Revolution, in the time of
the Enlightenment. La Mettrie is in trouble because of other provocations he published.
His books are burned, and he is living in exile.
In L’Homme Machine, La Mettrie undertakes for the first time the radical attempt
to reduce the higher human functions to bodily roots, even to simple mechanical
explanations. This essay cannot be the place to contribute to the ongoing and,
perhaps, never ending discourse about the machinic component in humans. It has
been demonstrated often enough that we may describe certain features of human
18I only have a German edition. The text can easily be found in libraries.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
92 F. Nake
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
behaviour in terms of machines. Although this is helpful at times, I do not see any
reason to set both equal.
We all seem to have some sort of experienced understanding of construction and
intuition. When working and teaching at the Bauhaus, Paul Klee observed and noted
that “We construct and construct, but intuition still remains a good thing.”19 We may
see construction as that kind of human activity where we are pretty sure of the next
steps and procedures. Intuition may be a name for an aspect of human activity about
which we are not so sure.
Construction, we may be inclined to say, can systematically be controlled; intuition,
in comparison, emerges and happens in uncontrolled ways. Construction
stands for the systematic aspects of work we do; intuition for the immediate, nonconsiderate,
and spontaneous. Both are important and necessary for creation. If Paul
Klee saw the two in negative opposition to each other, he was making a valid point,
but from our present perspective, he was slightly wrong. Construction and intuition
constitute the dialectics of creation. Whatever the unknown may be that we call
intuition, the computer’s part in a creative process can only be in the realm of construction.
In the intuitive capacities of our work, we are left alone. There we seem
to be at home. When we follow intuitive modes of acting, we stay with ourselves,
implicit, we do not leave for the other, the explicit.
So at the end of this mental journey through the algorithmic revolution (Peter
Weibel’s term) in the arts, the dialectic nature of everything we do re-assures itself.
If there is anything like an intuitively secure feeling, it is romantic. It seems essential
for creativity.
In the first narration, I presented the dense moment in Stuttgart on the 5th of
February, 1965, when computer art was shown publicly for the first time. If you
tell me explicitly, Georg Nees told the artist who had asked him—if you tell me
explicitly how you paint, then I can write a program that does it. This answer concentrated
in a nutshell, I believe, the entire relation between computers, humans,
and creativity.
The moment an artist accepts the effort of describing how he works, he reduces
his way of working to that description. He strips it of its embedding into a living
body and being. The description will no longer be what the artist does, and how
he does it. It will take on its separate, objectified existence. We should assume it is
a good description, a description of such high quality concerning its purpose that
no other artist has so far been able to give. It will take a lot of programming and
algorithmic skill before a program is finished that implements the artist’s rendition.
Nevertheless, the implementation will not be what the artist really does, and how he
does it. It will, by necessity, be only an approximation.
He will continue to work, he will go on living his life, things will change, he
will change. And even if they hire him as a permanent consultant for the job of his
own de-materialisation and mechanisation, there is no escape from the gap between
19(Klee 1928) Another translation into English is: “We construct and construct, but intuition is still
a good thing.”
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
3 Construction and Intuition 93
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
a human’s life and a machine’s simulation of it. Computers just don’t have bodies.
Hubert Dreyfus (1967) has told us long ago why this is an absolute boundary
between us and them.
The change in attitude that an artist must adapt to if he or she is using algorithms
and semiotic machines for his or her art is dramatic. It is much more than the cozy
word of “it is only a tool like a brush” suggests. It is characterised by explicitness,
computability, distance, decontextualising, semioticity. None of these changes is
by itself negative. To the contrary, the artist gains many potentials. His creative
capacities take on a new orientation exactly because he or she is using algorithms.
That’s all. The machine is important in this. But it is not creative.
The creation of a work that may become a work of art may be seen as changing
the state of some material in such a way that an idea or intent takes on shape.
The material sets its resistance against the artist’s will to form. Creativity in the
artistic domain is, therefore, determined by overcoming or breaking the material’s
resistance. If this is accepted, the question arises what, in the case of algorithmic
art, takes on the role of resistant material. This resistant material is clearly the algorithm.
It needs to be formed such that it is then ready to perform in the way the
artist wants it to do. So far is this material removed from what we usually accept
under the category of form, that it must be built up to its suitable form rather than
allow for something to be taken away. But the situation is similar to writing a text,
composing a piece of music, painting a canvas. The canvas, in our case, turns out to
be the operating system, and the supporting program libraries appear as the paints.
Acknowledgements My thanks go to the people who have worked with me on the compArt
project on early digital art and to the Rudolf Augstein Stiftung who have supported this work
generously. I have never had such wonderful and careful editors as Jon McCormack and Mark
d’Inverno. They have turned my sort of English into a form that permits reading. I also received
comments and suggestions of top quality by the anonymous reviewers. All this has made work on
this chapter a great and enjoyable experience.
References
Anon (1965). Bald krumme linien. In Der Spiegel (pp. 151–152).
Bense, M. (1965). Aesthetica. Einführung in die neue Ästhetik. Baden-Baden: Agis. This is a collated
edition of four books on aesthetics that appeared between 1954 and 1960. Aesthetica has
been translated into French and some other languages.
Birkhoff, G. (1931). A mathematical approach to aesthetics. In Collected mathematical papers
(Vol. 3, pp. 320–333). New York: Am. Math. Soc.
Cohen, H. (2007). Forty-five years later. . . . http://www.sandiego.gov/public-library/pdf/
cohencatalogessay.pdf.
Dreyfus, H. (1967). Why computers must have bodies in order to be intelligent. The Review of
Metaphysics, 21, 13–32.
Duchamp, M. (1959). The creative act. In R. Lebel (Ed.), Marcel Duchamp (pp. 77–78). New York:
Paragraphic Books.
Frank, H. (1964). Kybernetische Analysen subjektiver Sachverhalte. Quickborn: Schnelle.
Gerstner, K. (1963). Programme entwerfen. Teufen: Arthur Niggli. Second ed. 1968, third ed. 2007
in English under the title Designing programmes. Baden: Lars Müller.
EDITOR’SPROOF
Book ID: 303794_1_En, Date: 2012-07-30, Proof No: 1, UNCORRECTED PROOF
94 F. Nake
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
Glowski, J. M. (Ed.) (2006). Charles A. Csuri: beyond boundaries, 1963-present. Columbus: Ohio
State University.
Guilford, J. P. (1950). Creativity. American Psychologist, 5, 444–454.
Gunzenhäuser, R. (1962). Ästhetisches Maß und ästhetische Information, Quickborn: Schnelle.
Hentig, H. v. (1998). Kreativität. Hohe Erwartungen an einen schwachen Begriff. München: Carl
Hanser.
Herzogenrath, W., & Nierhoff, B. (Eds.) (2006). Vera Molnar. Monotonie, symétrie, surprise. Bremen:
Kunsthalle. German and English.
Herzogenrath, W., Nierhoff, B., & Lähnemann, I. (Eds.) (2007). Manfred Mohr. Broken symmetry.
Bremen: Kunsthalle. German and English.
Hollinger, L. (Ed.) (1999). Vera Molnar. Inventar 1946–1999. Ladenburg: Preysing Verlag.
Keiner, M., Kurtz, T., & Nadin, M. (Eds.) (1994). Manfred Mohr. Weiningen-Zürich: Waser Verlag.
German and English.
Kittler, F. (1985). Aufschreibesysteme 1800/1900. München: Fink. English: Kittler, F. (1990). Discourse
networks 1800/1900. Stanford, with a foreword by David E. Wellbery.
Klee, P. (1928). Exakte versuche im bereich der kunst.
Knuth, D. E. (1968). The art of computer programming. Reading: Addison-Wesley. Planned for
seven volumes of which three appeared from 1968 to 1973. Resumed publication with part of
Vol. 4 in 2005.
Lunenfeld, P. (1999). The digital dialectic. New essays on new media. Cambridge: MIT Press.
McCorduck, P. (1990). AARON’s code: meta-art, artificial intelligence, and the work of Harold
Cohen. New York: Freeman.
Moles, A. A. (1968). Information theory and esthetic perception. Urbana: University of Illinois
Press. French original 1958.
Nadin, M. (2011). Semiotic machine. An entry of the Semiotics Encyclopedia Online. http://www.
semioticon.com/seo/S/semiotic_machine.html.
Nake, F. (1974). Ästhetik als Informationsverarbeitung. Vienna: Springer.
Nake, F. (2009). The semiotic engine. Notes on the history of algorithmic images in Europe. Art
Journal, 68, 76–89.
Nees, G., & Bense, M. (1965). Computer-grafik (19th ed.) Stuttgart: Walther.
Nöth, W. (2002). Semiotic machines. Cybernetics & Human Knowing, 9, 5–21.
Shannon, C. E., & Weaver, W. (1963). The mathematical theory of communication. Chicago: University
of Illinois Press.
Stern, W. (1912). The psychological methods of intelligence testing. Baltimore: Warwick and York.
Transl. from the German.
Sundin, B. (Ed.) (1980). Is the computer a tool? Stockholm: Almqvist & Wiksell.
View publication statsView publication stats