Dr. Ron Eglash
Published in Cultural Studies 12:382-409, 1998
Cybernetics and American Youth Subculture
Cybernetics, the
interdisciplinary application and synthesis of information sciences, has
occupied a central place in analyses of postmodernity as a cultural condition.
Lyotard (1984:60) contends that a focus on information has led to sciences
based on “undecidables, the limits of precise control, conflicts characterized
by incomplete information, ‘fracta,’ catastrophes, and pragmatic
paradoxies,” an anarchic freeing of information in which he sees socially
liberating potential. Jameson (1991) sees the postmodern theme of
“depthlessness” as both the product of cybernetic simulacra and the symptom of
a social loss of history and place. Poster (1990) and Turkle (1995) suggest
that cybernetic modes of communication have resulted in postmodern
subjectivities in which personal (and even national) identity is unstable,
multiple, and diffuse. Martin (1994) weaves the narratives of flexible
accumulation in capitalism with the flexibility fostered by cybernetic models
of the immune system. Haraway (1997:219-229) produces an 11 page taxonomy of
intersecting cultural, natural, and technological categories offering vivid
illustrations of a three-stage mutual construction between postmodern culture
and cybernetics.
These interpretative
accounts are evocative, literary, and speak to the need for an emancipatory
poetics that can decode and recode the otherwise ineffable dynamics of a
threatening technopolitical economics. They stand in sharp contrast to the
kinds of “internalist” historical narratives that much of the scientific
community has produced to account for these technological changes. Take, for
example, the internalist explanations for the rise of chaos theory.
Mathematician John Franks (1989:65) put this baldly in his negative review of
Gleick's famous popular text Chaos -- making a new science, where he
insisted that Gleick had missed "the obvious explanation of the sudden
interest in chaos in many branches of science," which was "the advent
of inexpensive, easy-to-use digital computers." In a rebuttal to Franks'
review, Gleick did not disagree with the explanation, but quoted his own
perpetual and insistent observations on the importance of computers, which Franks oddly seemed to have overlooked
(and one which he apologized for in his reply).
This explanation of technological determinism is quite common in both
popular and professional circles. For
example, it was included in Heinz-Otto Peitgen's introduction to his fractal
geometry course at UCSC, in Stewart (pg 138), in Pagels 1988 (pg 85), and even
by Benoit Mandelbrot (1989 pg 6). There is no mention of any social changes in
connection to this technological transition in these accounts.
The two approaches I
have outlined here -- interpretative and internalist -- are complimentary in
their abilities and blind spots. On the one hand, interpretative evocations can
easily lose touch with the internal logic of a particular science. As an anthropologist
I am delighted to see Haraway write “transuranic elements:transgenic organisms
:: the cold war:the new world order,” but as an engineer it leaves me
bewildered; there is no internal connection between transuranic and transgenic.
The same could be said of the conflation of uncertainty in quantum mechanics
and uncertainty in deterministic chaos in Lyotard (1984) and Hayles (1990), or
Jameson’s comparison of cybernetic and architectural depthlessness. But this
ability to make leaps between internally separated technological features is
crucial to the power of these accounts in revealing the elusive yet systemic
relations between science and culture. Conversely, the internalist view of
steady progress through causal chains of scientific production, while providing
a high-fidelity representation of the technical features, becomes blinded to
their deeply intertwined social patterns.
This essay will attempt
to combine the technologically centered view of internalist accounts of
cybernetic sciences with the cultural semeiotics of interpretative
representation. As in Haraway, the analysis will distinguish between three
historical phases: 1) modern cybernetics, focused on digital hierarchical
systems, which reached its high point in the late 1960s; (2) transitional
postmodern cybernetics, focused on analog decentralized systems, which started
in the late 1970s; and (3) stable postmodern cybernetics, based on a synthesis
between the analog/digital and centralized/decentralized oppositions, which
started in the late 1980s. The technical
characterization of this sequence has been remarked upon by some of the
participating scientists and engineers, who see it as a “natural” chain of
events. What does beg for an
explanation, however, is the way in which youth subculture closely parallels
these changes. The simultaneous nature of the parallel transitions suggests
that there are causal links between cybernetics and popular culture which work
in both directions.
1) Information and
representation in cybernetics
Cybernetic theory is
based on two dimensions of communication systems (figure 1). One is the information structure, the other
is the physical
representation of that information. The most fundamental characterization for an
information structure is its computational complexity, which is a measure of
its capacity for recursion. The extremes
of this measure give us the distinction between a nonlinear, self-organizing
system, in which information is recursively assembled in a bottom-up structure,
and a linear hierarchy, in which information is imposed top-down.
The most fundamental
characterization for representation is the analog-digital distinction. Digital representation requires a code table
(the dictionary, the Morse code, the genetic code, etc.) based on physically
arbitrary symbols (text, numbers, flag colors, etc.). Sassure specified this characteristic when he
spoke of the "arbitrariness of the linguistic signifier." Analog
representation is based on a proportionality between physical changes in a
signal (e.g. waveforms, images, vocal intonation) and changes in the
information it represents. A computer
mouse, for example, guides cursor movement in proportion to hand movement
(although there is lots of digital circuitry mediating in-between). The difference between analog and digital is
often referred to as a dichotomy between continuous and discrete signals,
although that is quite misleading, since an analog parameter might change only
in discrete increments. It is more
important to consider that digital systems use grammars, syntax, and other
relations of symbolic logic, while analog systems are based on physical dynamics
-- the realm of feedback, hysteresis, and resonance (Dewdney 1985, Eglash
1993).
This essay will examine
these two dimensions of cybernetics at three different historical moments; I
have arbitrarily used the years 1967, 1977, and 1987 as foci.[1] As noted in the first paragraph, these
moments can be associated with (although certainly not reduced to) changes in
technological emphasis on the two cybernetic dimensions, from the promotion of
digital hierarchies typical of 1967, to the new focus on analog decentralized
systems starting around 1977, to the synthesis of these approaches starting
around 1987. These technical changes
will be illustrated by examining instances of the use of these two cybernetic
dimensions in four different disciplines:
machine intelligence, visual neurobiology, cerebral lateralization, and
mathematical modelling.
2) Cybernetics in the
modern era
It was... a considerable
revelation when writing came to detribalize and to individualize man....
Cybernation seems to be taking us out of the... world of classified data and
back into the tribal world of integral patterns and corporate awareness
(McLuhan 1966 pg 102).
In the first few years
of American cybernetics, analog and digital systems were seen as
epistemologically equivalent; both considered capable of complex kinds of
representation. In 1952, for example, a
study by the
While artificial brains
were being centralized, the natural brain was also losing its distributed
characterization. In Hubel and Wiesel's
model of visual neurobiology, each cell of the retina used lateral inhibition
to break the visual field into discrete points, like a digital video. These neurons were connected in rows to cells
in the first cortical layer, so that each first layer cell would fire only if
there was a line in the visual field that corresponded to its particular
orientation. These were connected to the
next cortical layer so that its cells could detect edges, corners and the like,
and so on to successive cortical layers in an increasing hierarchy of
specificity. Following this to its
logical conclusion resulted in paradox, because the cells at the top layer
would have to be so specific that there would be one cell for every image you
can recognize (often referred to as the "grandmother cell" for its
hypothetical ability to fire only when the visual field held an image of your
grandmother), but Hubel and Wiesel had excellent experimental evidence for the
lower layers, and the digital video metaphor was easy to understand.
At a much larger scale,
the models for cerebral lateralization in the late 1960s not only spanned the
whole brain, but reached across human behaviors from motor movement to gender
(Star 1979). The left hemisphere was
seen as digital, reductionistic, mathematical, and male; the right was analog,
holistic, artistic, and female. The
association was not necessarily sexist per se; Bogen (1969) for example defends
the right hemisphere as unjustly disparaged, and suggests that since these same
attributes (intuition, gestalt perception) are attributed to women, the
belittlement has a sexist bias.
It may seem all too easy
to make social-historical characterizations of such wildly generalizing claims
from behavioral biology, but even in mathematics, at the opposite end of the
hard science/soft science spectrum, the same bias towards digital linear
hierarchies can be seen during this time period. As noted in Gleick's (1987) historical
account, linear approaches to mathematical modelling, such as piece-wise linear
approximation in dynamical systems, linear programming theory, and symbolic
logic were well-funded, while nonlinear dynamics received little
attention. The foundations for what is
today known as deterministic chaos -- an explicit illustration of the
computational complexity of analog recursion -- were largely ignored. Physicist Joseph Ford, for example, recalled
his early failure to introduce notions of complex nonlinear behavior in a
lecture on the Duffing equation:
When I said that?
Jee-sus Christ, the audience began to bounce up and down. It was "My daddy played with the Duffing
equation, and my granddaddy played with the Duffing equation, and nobody seen
anything like what you're talking about."
You would really run across resistance to the notion.... What I didn't
understand was the hostility (Gleick pg 305).
Meteorologist Edward
Lorenz met similar resistance: his seminal paper on deterministic aperiodicity
was ignored despite the efforts of respected supporters such as Edward Spiegal
and James Yorke. As noted earlier, most mathematicians have insisted that the
change was due to the availability of digital computers. But the implication
that digital computers can explain the revolution in nonlinear dynamics is
misleading. The emergence of chaos as a legitimate and valuable research topic
in the early 1970s was often carried out on analog computers. This was
true for Ueda in
Rather than the lack of
digital computers, a better explanation for the dominance of linear
hierarchical modelling in the 1960s can be found in the political affiliations
of the scientists involved. Ralph Abraham, for example, who would later become
the main mathematical supporter for Shaw’s chaos cabal at UCSC, reports that
the political lines in the mathematics department at UC Berkeley in the 1960s
divided right versus left along linear versus nonlinear (personal communication
1991). James Yorke, whose attempt to promote the work of Lorenz was met with
staunch resistance, was also known for his war resistance, having made direct
use of his mathematical skill during the antiwar demonstrations in his analysis
of a fraudulent government photo. In 1966 mathematician Steve Smale learned
that his research funding for nonlinear dynamics was being revoked by the
NSF. But at the time Smale was also
subject to a subpoena by the House Un-American Activities Committee due to his
leftist activism; it was not clear whether his offense was non-linear science
or un-American politics.
This political
polarization along both the analog-digital and recursive-nonrecursive
dimensions was wide-spread in the late 1960s, and this was nowhere more evident
than in cybernetics itself. The process
by which the analog-digital dichotomy was established as field for
contestations began at the origins of cybernetics in the first Macy conferences
of the late 1940s. This was commented on
at the time by Gregory Bateson.
There is a historic
point that perhaps should be brought up; namely, that the
continuous-discontinuous variable has appeared in many other places. I spent my childhood in an atmosphere of
genetics in which to believe in 'continuous' variations was immoral. ...There is a loading of affect around this
dichotomy which is worth our considering.
(von Foerster 1952)
Bateson eventually took
a position opposite to that of his father's; for him analog systems were closer
to the Real, and thus preferred as a more holistic or Natural form of communication.
Bateson, together with Margaret Mead, came to represent the anthropological
sciences for cybernetics. Mead took advantage of this position to offer some
recursive perspectives on cybernetics at the first Annual Symposium of the
American Society for Cybernetics in 1962.
She began her introductory speech, "Cybernetics of
Cybernetics," with some memories of the foundational meeting of the
society for General Systems Theory.
I suggested that,
instead of founding just another society, they give a little thought to how
they could use their theory to predict the kind and size of society they
wanted, what its laws of growth and articulation with other parts of the
scientific community should be. I was
slapped down without mercy. Of all the
silly ideas, to apply the ideas on the basis of which the society was being
formed to Itself! (von Foerster et al 1968 pg 10).
Mead was already famous
for her blending of scientific analysis with cultural critique, dating from her
refutations of theories on the biological determination of gender
characteristics in the 1920s. In this talk she warned against the unreflective
use of cybernetics, pointing to L.F. Richardson's pacifist approach to
mathematical models of warfare, and Edmund Bacon's use of cybernetics in participatory
city planning as alternatives. Mead's
articulation of the ironic lack of "circular self-corrective systems"
in the science of circular self-corrective systems underscored the impending
politicization of recursion in cybernetics.
This association of
analog representation with the Real and recursion with liberation was taken up
by many leaders of the cybernetics community: Kenneth Boulding (a founder of
General Systems Theory), Tolly Holt (director of Advance Systems in Princeton),
C.H. Waddington in biology, Karl Pribram in neuropsychology, Hazel Henderson in
social systems, and Magoroh Maruyama and Heinz von Forester in cybernetics
proper, to name a few. The humanist
translation of recursion (in his terminology "feedback") as
liberation was previously proposed by cybernetics founder Norbert Wiener (Heims
1984), and the view of analog systems as closer to the Natural or the Real was
independently championed by Francisco Varela[2] and Humberto Maturana,
who worked under Perceptron founder Warren McCulloch.
3) Youth subculture
and cybernetics in the modern era
Youth subculture studies
(e.g. Hebdige 1979,1988, McRobbie 1994,
Hall 1980) have made the reading of stylistic expressions (music, dress,
speech, etc.) in these social groups accessible to semiotic analysis, and this
is nowhere more evident than in the contrast between the youth subculture and
parent culture of the 1960s. The contrast between the recursive analog dynamics
in the counter culture and digital hierarchy in the “military-industrial
establishment” was evident in the geometry of material designs chosen by each
subculture. The psychedelic aesthetic of hippy adornment was essentially what
we would now call fractal: paisleys with-in paisleys with-in paisleys, flowing
robes with folds of folds, rippling organic hairstyles. Establishment icons
were in the Euclidean shapes of rectangles, circles and triangles, as evidenced
by the computer font introduced with the IBM punched card system. It is no
coincidence that hippies accused “straights” of being “square.”
That the (primarily
white[3]) romantic organicism of
American youth subculture in the 1960s would be taken up by a scientific
community is not surprising given the personal ties of these individuals. Kenneth Boulding's house, for example, was a
central location for organizational meetings of the Students for a Democratic
Society (SDS) in the early 1960s (Kerman 1974, Wright 1989). Many of the first SDS manifestos were based
on the idea of "face to face communication" as a more ethical means
of representation than digital encodings; this analog anarchism came from Paul
Goodman, who was in turn influenced by a line of holistic scientists from Peter
Kropotkin to Gregory Bateson (Goodman's next-door neighbor during his stay in
The conservative half of
this cybernetic politics, starting with von Neumann in the 1940s, is already
well-described in a literature which, in my view, over-emphasizes this side due
to technophobic bias (e.g. Lilienfeld 1976). Here we see rationalists with an
affirmation of the artificial, and an authoritarian love for imposed
control. The digital hierarchists
managed to take control of the majority of funding and institutions for
cybernetic research during the 1960s, and made an effort to suppress the
legitimacy, if not the existence, of the analog decentralists. It is a sad irony that critics such as Lilienfeld
have simply duplicated that erasure.
 |
 |
This portrait of the
historical changes in the cybernetic map, from its flat equipotential surface
of 1947 to its bifurcated basins of attraction in 1967, should be seen as
coupled to an similarly changing map of politics, as I have drawn for Bateson
and Mead (figure 3).
Here the changes
are coming from opposite causal directions for the two individuals. Mead's political involvement was a long-term
commitment for her; but the technical dimensions were not polarized for Mead
until after she understood how the politics might couple with this cybernetic
map. For Bateson, the direction of
causality was the reverse. His collaborator
Ken Norris (personal communication) suggests that for Bateson, cybernetics was
an opening into the political arenas he had previously avoided, and Bateson's
daughter has made similar implications (M.C. Bateson 1984 pp 217-220). Bateson was quite fixed on the technical
significance of both analog representation and recursion, but his political map
was flat until he saw how a hyperbolic humanism could match his own cybernetic
topology.
The actor-network
formalism of Latour, Callon and others is typically portrayed as a struggle in
attempts to recruit scientists (and in more recent versions, non-human actors)
into a web of power relations that supports a particular device or theory. But
the polarizations in these two cybernetic dimensions goes far beyond a
particular laboratory or even a single discipline, and have specific links to
socio-political positions of their day. Between the decentralized organicism of
hippie lifestyle and the hierarchical linearity of industrial and governmental
institutions, the 1960s could be summarized as a fight between the natural
chaos of counter-culture and the artificial order of "the
establishment."
4) Transitional
postmodernism and youth subculture
To this career I did
ordain
Doin' things the
anthropologists can't explain
Arrhythmic sounds float
like
Altering any
scientifical theory
Sweet Tee, "On the Smooth Tip"
African-American urban
youth movements of the transitional postmodern era, often grouped under the
term Hip-Hop, constitute yet another move in the analog-digital dualism game;
but this time with revolt on the digital side.
The term "digital" is itself used by these movements (best
exemplified by the popular rap group "Digital Underground") to
signify a guerilla cybernetics, both in terms of street-wise technical
competence and in cultural communication forms (Rose 1994, Dery 1994, Eglash
1995).
Rap is notorious for its
non-melodic sound: a disjoint collage of flat punctuated speech, spliced sound
bytes from James Brown, The Andy Griffith Show, video game lasers, and
above all the scratch: deejays playing the normally silent back-cue of the
record in time to the beat, letting the raw signal of a misused stylus
interface the acoustic codes into a single mutant form. Although not a commercial success until the
Sugar Hill Gang's "Rapper's Delight" of 1979 and Grandmaster Flash's
"The Message" of 1982, by 1977 most of these forms were already
existing from low-tech necessity: budget disco parties without the money for a
band or the passivity for records. Out
of the pit of commercialized pop music -- the gold-chained discotheque scene of
John Travolta and Studio 54 -- scratch turned tables on the turntables.
At the same time rap was
cutting organic sounds into oppositional symbolics, breakdancing was digitizing
the human body. Like rap, histories of breakdancing (DiLorenzo 1985, Hager
1984) cite traditions from African prehistory to contemporary African-American
dance, in which discrete moves, rapid-fire symbolic gestures and percussive
"popping" originated. Breakdancing also uses a stylistic collage of
odd elements: "King Tut" friezes, Russian folk dance and street
gymnastics are fused with the pro-artifice titles of "Electric
Boogie," "Moonwalk" and "Robot." The android movements of Shields and Yarnell,
a husband and wife performance team seen widely on popular television in 1977,
were often incorporated by breakdancers in routines which code the
machine-human interface in much stronger ambiguity than anything seen in the
modern era. While an element of irony
and parody are always present, there is also pride in the technical competence
required to take on the cybernetic stronghold of industry, and a historical
claim of ownership for these powerful modes of representation.
A third important
element is graffiti: not only a claim to ownership of coding, but coding as a
claim to ownership. In 1969 an
unemployed 17-year-old New Yorker started writing his nickname together with
his street number on public spaces; soon Taki 183 was joined by hundreds of
male and female writers of various ethnicities, genders and economic
backgrounds. Since the majority are
poor, the activity is often interpreted as a rebellion against their economic
trap: a signature marks possession.
Castleman (1982), in a
solid ethnographic study of
All three of these
aspects of hip-hop include instances of analog communication, but all three
emphasize the digital features in ways which make the movement as a whole far
more friendly to artificial states of being: to urban life, machines, science,
and even to artificial origins.
Post-liberated black
hair-styling emphasizes a 'pick n' mix' approach to aesthetic production,
suggesting a different attitude to the past in its reckoning with
modernity. The philly-cut on the
hip-hop/go-go scene etches diagonalized lines across the head, refashioning a
style from the 1940s where a parting would be shaved into the hair (Mercer
1988, pg 51).
The way in which the use
of arbitrary symbols signifies artificiality is a crucial part of this
postmodern distinction. The
The hums, grunts, and glottal attacks of central
While Black youth
subculture was reinventing itself as digital, white youths made a similar
transformation. In 1977 - at the all-time high of teenage suicide in the
The arbitrariness of
symbols and symbolic arrays is essential to all punk material culture. The clothing is a an urban screech of plastic
and metal; the hair a neon dyed, machine sculpted artifact; the make-up flaunts
the taboos of all fashion advice; and the music itself is an unrelenting
succession of noise. Despite the
frequent use of the term "chaos" to describe punk (eg Johnny Rotten's
famous line "we're not into music, we're into chaos"), the punk
aesthetic bears no resemblance to the hippy use of natural flowing chaos, to
patterned organic shapes which opposed the establishment's Euclidian
order. The disorder of punk is 'white
noise,' the randomness of disrupted syntax, "a place where the social code
is destroyed and renewed" (Kristeva, 1975). The songs of the Minute Men
are all under 60 seconds, the Circle Jerks's songs not only standardized in
time, but repeat the same chord pattern
as well. Just as Rap uses speech to
break up the organic flow of music, punk interjects order where it does not
belong. In hippy culture plastic meant the establishment's order (as in
"plastic hippy," or Dustin Hoffman's economic advice in The
Graduate). To punks plastic is
"plastic" in the sense of transformable; its excesses could force the
system against itself, drowning institutional order in the noise of its own
symbolics.
If punk is so intent on
self-defined symbolics, doesn't that mean they subscribe to a modernist program
of recursive control or humanist self-definition? Far from it. In terms of individuals, humanism
differentiates between control from false authority and control from the true,
authentic self. With no interest in authenticity or the "true self,"
Punks promote fakes as the only legitimate sincerity. In terms of collective activity, the same
disruption occurs. Chambers (1985)
describes this for punk efforts to subvert the recording industry.
Punk's rough populism,
what it would call... 'street level' music, was frequently translated into a
flood of small, independent record labels.
These apparently mushroomed around the subculture's early leitmotif: musical
populism + recording independence = cultural autonomy. But as the idea of
'authenticity' (hence the measure of 'autonomy') was also a rather non-punk concept, the whole
formula remained precarious.(pg 228).
In contrast to the 60s
counter-culture's search for recursive control, the postmodern youth culture
fought for a critique of this humanist self-reference. Just as the paradox of Rap's natural artifice
opened a space for new cultural constructions, punk's recursive critique of
recursion provided another opportunity
for creating new kinds of resistance. Just as the counter-culture of the 1960s
co-evolved with
cybernetics, defining itself as the cybernetic opposite of digital hierarchy
(in parallel with the military, industrial and institutional move to define
itself as the cybernetic opposite of analog decentralization), the affirmation
of digital representation and skepticism about recursive control in postmodern
youth subcultures occurred in tandem with a scientific revolution
in the opposite
direction. Postmodern cybernetics
requires analog recursion; it is here that science seeks chaos.
5) Cybernetics in transitional postmodernism
Despite the
institutionalization of digital representation and linear hierarchy in many
areas of modernist technoscience, 1967 -- the "summer of love" -- was
the peak for cybernetic order, and by 1977 this had begun to reverse. In the same year that the Sex Pistols
declared their love for the sharp Euclidian lines of urban artificiality,
cybernetic models of the military-industrial establishment began to side with
the natural holistic flow of chaos.
Of course, chaos theory
itself is the clearest example of this transition. Fractal geometry, dynamical
systems theory, cellular automata, and other models of self-organizing process
created a dramatic acceleration in the ability of scientists to study the
recursive flow of information which underlies a wide variety of complex
nonlinear phenomena. As noted earlier, it was by reinterpreting physical
dynamics as analog representation, not by sticking to the old guns of progress
in digital computers, that the chaos revolution made its real impact.
It is here that the
semiotic claim for a special status of digital representation is proved wrong:
analog waveforms can express recursive complexity as well as any symbolic code
can. But many scientists during this
transitional moment went entirely overboard, now claiming that analog systems
were computationally superior to digital. Dewdney (1985), for example, announced that
he had devised an analog version of the Turing machine,[4] and had proved that it
could out-perform the original digital Turing machine (I later wrote to ask him
for details, but he replied that those notes were "buried somewhere"
in his files). Researchers at the
Princeton University Computer Science Dept (Vergis et al 1985) claimed to have
an analog machine that could solve an NP-complete problem in less than
exponential time -- a performance which is theoretically impossible for any
digital machine. Their claims were
disproved in Rubel (1989). A number of
prominent physicists (Pagels, Feynman, Penrose and others) took refuge in quantum indeterminacy to claim
that there are analog systems (i.e. the human brain) that have computational
powers superior to digital Turing machines.
The study of non-deterministic Turing machines is decades old, and has
not demonstrated that there is a computational advantage to these systems.[5]
In machine intelligence,
both hardware and software went through a sudden transition away from
centralized hierarchies. In the digital
parallel processing approach, many von Neumann-type (linear centralized)
processors are linked together, working on different parts of the problem at
the same time. The problem can be
divided up in different ways with varying degrees of autonomy. At the most extreme decentralization we
return to the perceptron. It turned out
that Minsky and Papert's analysis did not take into account the possibility of
perceptrons in which there was a "learning layer" mediating between
input and output; this minor addition resolves the pattern recognition
limitations they had raised (Pollack 1989).
In the simulated
annealing approach (cf Kirkpatrick et al 1983), an optimal state for an
information system is approached by allowing carefully controlled randomness to
perturb local elements, tossing them into desired configurations (i.e.,
"hill-climbing" to local maxima) and then decreasing the randomness
before they are tossed out. This method makes use of the digital simulation of
a physical process (thermal control of annealing in metals). The holistic extreme for hardware was the
coupled oscillator approach (cf Hopfield 1982) where computation was carried
out by decentralized networks of analog units modeled on neurons.
The software
decentralization in machine intelligence research was first utilized in the
language ABSET (Elcock, McGregor and Murray 1971), which allowed data-directed
control. More explicit decentralization
was achieved in the "actor" formalism of the work of Hewett (1977)
and the knowledge representation system of Bobrow and Winograd (1977). In
explaining decentralization in AI software, Pylyshyn (1981) declares that
"this means loosening the hierarchical authority relation common to most
programs, and distributing authority in a more democratic manner, allowing for
'local initiative'" (pg 79). By
the 1980s even Minsky suggested that decentralization was a good idea in software
(and in his 1986 Society of Mind he claimed that's what he had been
thinking about all along).
In visual neurobiology
the challenge to Hubel and Weisel's work began with Campbell and Robinson
(1968), in which suggestive evidence was presented for global frequency
functions (e.g. the fourier transform) in the striate cortex. For example, one experiment showed a neuron
which fired optimally for an image of a monkey's hand: a specificity which is
too great for Hubel and Weisel's model, but fits well with the global frequency
model since the monkey's fingers form a spatial grating. However, it was not until DeValois et al
(1978) that "critical" experiments which purported to decide the
issue actually challenged the dominance of local feature extraction
models. Many researchers concluded that
the best analogy for visual neurobiology was not the digital video camera, but
rather the analog medium of a laser hologram.
As in visual
neurobiology, changes in cerebral lateralization models were presented as
compelling objective data requiring a different conceptual frame, not as a more
sensible or fashionable theory. Once
researchers had settled on a postmodern reversal -- assigning the digital left
hemisphere to women and the analog right hemisphere to men -- a
multidisciplinary complex of theoretical apparatus was quickly assembled. Levy (1978) developed an origin story in
which "man the hunter" needed the gestalt skills of mapping and spear
tossing while "woman the gatherer and childrearer" needed her social
communication skills. While Buffery and
Gray (1972, quoted in Star pg 68) suggested that female superiority in symbolic
systems corresponded with "good performance on clerical tasks," Geschwind
(reported in Kolata 1983) linked a supposed innate mathematical superiority in
males to testosterone effects in fetal brain development, which would then
produce "superior right hemisphere talents, such as artistic, musical, or
mathematical talent" (pg 1312).
Note that mathematics in this description has moved from its 60s
association with the symbolic left hemisphere to a postmodern association with
the analog right hemisphere.
In summary: while the
60s counter-culture sought analog naturalism and recursive humanism,
anti-hegemonic youth subculture in transitional postmodernism fought for a
legitimation of artifice and a critique of humanist self-reference. At the same time, mainstream information
sciences [6] switched to a serious
consideration of analog representation and nonlinear decentralization.
6) Stable
postmodernism and cybernetics
Starting around 1987,
cybernetic models again shifted; this time to a synthesis of the competing
opposites. The term "stable
postmodernism" is appropriate in that the move resolves both technical and
social tensions of the previous sharp reversal in these information dimensions.
When we last left
mathematics, it had moved from the Euclidian linearity of the 1960s to the
fractals and dynamical chaos of the 1970s.
In the late 1980s these two approaches were brought together. This was nicely illustrated by the
"multifractal" approach initiated by Uriel Frisch (Nice Observatory)
and Giorgio Parisi (
In machine intelligence,
we saw the postmodern transition from linear digital hierarchy in the 1960s to
both decentralization and analog systems in the late 1970s. By 1987 many researchers decided that the
difficulty of nonlinearity, despite the new knowledge rising from chaos theory,
was still too great for a pure analog approach, and that analog and digital
would need to work together in new hybrid machines. Marvin Minsky, for example, had trashed
holism in his modernist text Perceptrons, and had valorized it in his
transitional postmodernist text Society of Mind. In his stable postmodernism essay of 1988, he
suggested that the ultimate fulfillment of his "Society of Mind"
approach would be a synthesis of the two: analog neural nets as the individual
"agents" of his AI system, with centralized digital algorithms
providing its governing structure.
Pollack (1989)
differentiates between "the old new connectionism, circa
1980-1985" and "the new new connectionism" on this basis,
noting the crucial addition of digital representation to neural nets in
"semantically interpretable higher-order elements." The reverse adaptation can also be seen in
the work of Stephen Omohundro. In a
widely circulated technical report of 1987 he described some important
innovations for algorithms in digital centralized computers which could emulate
neural network behavior. Equivalences
between recursion in both analog and digital systems were also approached at
this time (eg Touretzky 1986), and later commercial production of the hardware
that could support such hybrids (Electronic Design 1989) also occurred.
Also appearing at this
time were neural nets based on a harmony theory similar to Bak's self-organized
criticality. Whereas modernist networks
were based on competition, and transitional postmodernist networks exploited
formal systems of cooperation, these stable postmodernist neural nets used a
balance of cooperation and competition to achieve their tasks (Maren 1990).
This change from the transitional trope of reversal to the stable trope of
synthesis or harmony allows a resolution of previous tensions, thus opening a
new acceptability. Harston and Maren
(1990) cite an exponential rise in the number of neural net projects starting
in 1987.
Analog-digital synthesis
in more banal computing machines had been suggested by the UCSC "Chaos
Cabal" in a 1981 NSF grant proposal, but the idea was premature and the
grant was rejected, although they did receive NSF support for work on their analog
computer. By the late 1980s NSF grant
directories specifically mentioned analog-digital hybrids as an area of
interest. Much of the theoretical work
in this new synthesis goes under the rubric of "computational dynamics,"
a title which not only emphasized the new fusion of analog and digital
computation, but could also serve as a respectable sounding alias for
"chaos."
Reassessment of the
analog optimism from transitional postmodern cybernetics was an important part
of this new effort toward harmony. While
many transitional postmodern researchers saw analog computation as uncharted
territory, a potentially utopic space where physical dynamics could beat a
Turing machine, the new stable postmodern theories viewed analog and digital as
equivalent representational forms. The
most formal recognition for this equivalence was in the work of Blum, Shub and
Smale (1989), which developed a system of mathematics that could show
limitations for analog computation which were isomorphic to the undecidability
limitations of Turing machines. Similar
implications were reached through work in symbolic logic by Rubel (1989), who
used this to critique the earlier claims of analog superiority by Vergis et al.
In visual neurobiology,
we moved from the 60s reductionist feature extraction hierarchy of Hubel and
Weisel to the 70s holistic fourier transform model of DeValois. But the experiments which had supported this
view of global spatial frequency detection did not necessarily negate Hubel and
Weisel's results, and following some debate a synthesis was proposed (MacKay
1981). The model finally adopted used
the Gabor transform, a combination of local and global information extraction
(figure 4) which had been developed in early communication theory, prior to the
polarizing split of the 1960s (Gabor 1946).
Empirical confirmation soon followed (Daugman 1984).
In cerebral
lateralization, modernist theories had suggested that the digital left
hemisphere was mathematical and male, and that the analog right hemisphere was
artistic and female. Transitional
postmodernism saw a reversal of this association, with mathematics as an
imagerial, holistic function of the right hemisphere and "man the hunter" preferring spatial
mapping in his neural evolution. In
stable postmodern lateralization, mathematical superiority is now said to
depend on the interaction between both hemispheres, again leaving room for male
superiority through studies reporting sex differences in interhemispheric
information transfer (first described in De Lacoste-Utamsing and Holloway
1982).
In all four disciplines,
information processing shifted from the holism reversal of the postmodern
transition to a new, more stable holism-reductionism synthesis. As in the case of the previous two
transitions, this paradigm not only extends across many sciences, but also in
popular culture. If hippies epitomize radical modernity, and punk and hip-hop
give us transitional postmodernism, then an obvious choice for the pop
subculture of stable postmodernism is New Age.
I hesitate to call this a youth subculture, since most of its members
are not very young, but in some ways it fetishizes youthfulness even more than
its predecessors.
6) New Age and the cult of harmony
Like hippy and punk rock
enthusiasts, New Agers are primarily white and middle class, and the semeiotics
of their subculture often concern this ethnic/class identity. Hippy signifiers could often be read as an
avoidance of whiteness, an attempted transformation into primitivised Native
Americans or orientalist holy beggars.
For hippies artificial (eg Euclidian shapes, synthetics) signified a
semantically empty whiteness, and natural (chaotic shapes, organic materials)
filled in this blank with a more earthy identity. Punk subculture claimed the blankness as its
own, using artificial and urban-associated materials to signify both their
acceptance of their ethnic identity and a refusal to ignore its ugliness. New Age members, like the cybernetic engineers
of their era, solve this problem of identity conflict neither by taking one
side or the other, but by harmonious synthesis.
The symbol which best
portrays this hybrid identity is the crystal embedded in unpolished rock, a
common motif in New Age fashion, art and literature. Like multifractals (and New Age hairstyles),
these objects are both Euclidian and fractal, both orderly and chaotic, and
carry the paradoxical message that something can be both natural and
artificial. In New Age subculture whiteness is just one of many ethnicities
-- one which happens to have an odd association
with artificiality, but that too is natural.
Hence the solution of identity conflict here allows New Age members to
"channel" through cultures (even those long gone) in ways which make
the most wanna-be of hippies look like white pride advocates.
The synthesis of
artificial and natural is also conveyed in New Age music, which typically uses
artificial (ie electronic) instruments to make natural sounds. But the natural sounds are themselves
supposed to reflect only the soothing harmonious essence of nature, the
"music of the spheres," and therefore have an ethereal, unworldly
character to them. Another popular
symbol among New Age participants is the yin-yang sign, which has a history of
orentalist interpretations that has shown great flexibility. In the 1960s the yin-yang was said to
represent holism, the true unity that lies beyond dualist illusions of day and
night, hot and cold, etc.. In New Age
subculture the same symbol is used to represent the unity of holism and
reductionism (which Toffler, writing primarily for a management science
audience, terms "wholism").
Figure 5 summarizes this comparison of youth subcultures and cybernetics
in the three historical moments outlined.
While this synthesis
solves many of the tensions that the transitional reversal posed -- for
example, justifying the lifestyles of thousands of ex-hippie computer
programmers -- it presents some conceptual problems. Should holism and reductionism be combined
holistically, or reductionistically? If
holism is not the answer at the first level, why should it be the meta-level
solution? Doesn't this imply two
different methods of synthesis, meta-holism and meta-reductionism, which could
again (perhaps in 1997) be brought together in yet another brilliant
philosophical coup?
My cynicism here comes
in part from an exasperation with the free-floating politics that this
metaholism has allowed. Riane Eisler's
1987 metaholism manifesto, The Chalice and the Blade, cites Karl Marx,
Donna Haraway, and Alvin Toffler as participants in her "glyanic"
(gynic/andro synthesis) wave, a sweeping inclusiveness that leaves one
wondering if there is anything she does not support. More ominous is the advocacy of prayer in
public schools suggested by metaholism theorists Charlene Spretnak and Fritjof
Capra. While this may indeed combine
holistic spirituality with reductionist institutionalization, it hardly seems
like the cutting edge of progressive politics that they claim.
As in the case of holism
pure and unmodified, I see no inherent ethical content to holism-reductionism
synthesis. Eisler, for example, suggests
that the concepts of both patriarchy and matriarchy imply domination, and
provides a utopic origin story in which the edenic state was a
"glyanic" holism-reductionism synthesis of "partnership"
rule. But it is hard to see how this
notion of partnership is inherently more liberating; it recalls Rich's
"compulsory heterosexuality" and bourgeois ideals of family norms
(see pg 39 of Eisler's text). And like the holism of transitional
postmodernity, this holism-reductionism synthesis is also hard at work in
military and industrial applications.
Robert Smithson of Lockheed, for example, recently suggested that the
ultimate defense computing system would "operate in a continuum of modes
ranging from analog to digital, and from network to sequential processor"
with an optimum combination sought for each problem (Stoddard 1990).
7) Conclusion
How can we account for
these parallel transitions between changes in information processing models and
changes in and popular culture? Hess (1995:18-19) categorizes two different
explanatory methods in science studies. In what he terms the “sociological approach,” empirically
identifiable social variables or actors are shown to have both correlation and
causal connection with particular technological counterparts. In contrast, Hess
refers to a “cultural approach” in which the correspondance between social and
technological features cannot be reduced to any particular factor or artefact.
Turkle (1995:264), for example, suggests that the correspondances between
postmodernity and cybernetics are cultural “deep assumptions” (here using a
phrase from Hayles) which “manifests
itself in one place and then another, here as developmental psychology and
there as a style of engineering, here as a description of our bodies and there
as a template for corporate organization, here as a way to build a computer
network and there as a manifesto of political ideals.”
Hess suggests that any
one researcher might switch between the cultural and sociological approach,
changing analytic tools as they see appropriate, but this makes the choice
sound like merely one of personal preference or rhetorical strategy. In the
discussion of relations between cybernetics and popular culture in the 1960s I
was able to give a good sociological account of the particular people, events,
and artifacts (or in Latour’s terms, the network actors) that created the
linkage. But this was much less apparent in the 1970s and 1980s. I did not decide
to switch explanatory methods; rather the nature of this particular relation
between the technological and the social has undergone an historic change.
There are two possible
explanations for such transformations. The first is that the technical-social
coupling only existed as a strong force during the 1960s, and that this was
like synchronizing two clocks -- what appeared to be coupling in later decades
was merely two independent dynamics with similar time constants. The second,
which I find more believable, is that the kinds of explicit, overt gestures of
political and technological allegiance that were appropriate in the 1960s would
have been overbearing in the 70s and 80s. The coupling between popular culture
and cybernetics was more subtle not because it was any less important, but
because the subtlety was itself historically conditioned.
One of the fiercest
battles of postmodernity has been the arguement between the “new left” of the
1960s and their postmodern counterparts. From the viewpoint of the sixties
activists, postmodernist analyses (i.e. poststructuralism) are merely an excuse
for navel-gazing (cf. Epstein 1995). Postmodernists generally reply that the
particular physical activity of massive street protests is not the only
possible theater for political movements, and that their political activism
takes more subtle routes. One of these has been a shift to new forms of
cultural representation, and it is here that we may find some of the mutual
influences between cybernetics and popular culture. Rose (1994: 62-96) makes a
detailed argument for the co-evolution of rap music and ceratin aspects of
digital sound engineering, and also notes (pg 22) that many of the computer
graphic effects that emerged in the later 1970s, such as morphing, were
foreshadowed by the fluid transformations of graffti and breakdancing from the
early 1970s. “In a simultaneous exchange, rap music has made its mark on
advanced technology, and technology has profoundly changed the sound of Black
music.”
Another example of this
subtle exchange occurs in the CALAB software documentation created by white
mathematician Rudy Rucker (1989). The
controls for a specific group of the cellular automata parameters are recalled
by analogy to rap artist names (pp 35, 37). When I asked him about the
reference, his reply was a striking mix of ethnic signifiers and science
fiction--a far more subtle connection than the romantic organicism that
characterized the 1960s:
Well, I just remember
thinking that it sounded like Black English, DB, DC, DD ... de brothers. I got
Run-DMC's record when I lived in
What stands out in this
history is not any clever end-game, but rather the strategic status of
oppositional moves. Norbert Weiner's
brief attempt to apply cybernetic recursion to labor unions in the 1950s was a
radical move, but in the 1980s worker "self-management" was more
about eliminating union activism.
Rucker's use of rap signifiers in his seminal cellular automata lab of
the late 1980s was only possible because of an politically unaccountable New
Age-style ethnic collage, but in the context of the typically all-white
signifiers of hacker technoculture (e.g. the MUD as multi-user dungeon, the
battle-ax as anti-virus icon, and other signifiers of a playful Tolkenesque
nordic mythology), it carries a cultural intervention that is not
insignificant. Anti-hegemonic critique
in cultural cybernetics works best when it does not depend on final answers,
unmovable grounds of truth, or ultimate formula. The only place to stand is to keep on
moving.
Bibliography
Adelman L, Zirk D, and
Lehner P. "Distributed Tactical
Decisionmaking." IEEE Syst Man
Cyber 16:6 pp 794-804 1986.
Alper, J.S. "Sex differences in brain
assymetry." Feminist Studies 11, 1 1985.
Bateson, M.C. Our Own Metaphor. NY: Knopf 1972.
Bateson, M.C. With a
Daughter's Eye. NY: Simon & Schuster, 1984.
Blum, L., Shub, M., and
Smale, S. "On a theory of
computation and complexity over the real numbers." Bull AMS, v21 N1
1-46, Jul 1989.
Bobrow, DG and Winograd,
T. "An overview of KRL." Cognitive Science 1 pp 3-46 (1977).
Bogen, JE Bulletin of
the
Boggs, J. "The Negro in Cybernation." in The
evolving society; the proceedings of the first annual Conference on the
Cybercultural Revolution - Cybernetics and Automation.
Callon, Michel, John
Law, and Arie Rip. Mapping the Dynamics of Science and Technology.
Campbell, FW and
Robinson, JG. "Applications of
fourier analysis to the visibility of gratings," J Physiol 197 pp
551-566.
Capra, F. The Tao of Physics.
Chambers,
Daft, R.L. and
MacIntosh, N.B. "A new approach to the design and use of management
information."
Daugman, JD "Two-dimensional spectral analysis of
cortical receptive field profiles," Vision Res 20: 847-856.
De Lacoste-Utamsing C
and Holloway RL (1982): "Sexual dimorphism in the corpus callosum." Science
216: 1431-1432.
Delany, Samuel. “Is
cyberpunk a good thing or a bad thing?” Mississippi Review 16 pp.28-35
1988
Dery, M. "Rap:
Taking it from the Streets. Keyboard Nov 1988.
Dery, M. "Black to
the Future: interviews with Samuel Delany, Greg Tate, and Tricia Rose.” in
Mark Dery (ed.) Flame
Wars,
De Valois RL, Albrecht
DG, Thorell LG (1978): "Cortical
cells: Bar detectors or spatial
frequency filters?" In Frontiers in Visual Science SJ Cool and EL Smith (ed.), Springer,
Dewdney, A.K.
"Analog Gadgets." Scientific American,
DiLorenzo, K.
"Breakdancing isn't New" The Crisis, pp 8-9,45. April 1985.
Eglash R "The
Cybernetics of Cetacea." Investigations on Cetacea, vol 16, 1984.
Eglash, R. "African influences in
cybernetics." in C. Gray (ed) The Cyborg Handbook.
Eglash, R. "Inferring representation type from the
fractal dimension of biological communication waveforms." Journal of
Social and Evolutionary Systems, 16(4):375-399, 1993.
Eisler, R. The
Chalice and the Blade.
Elcock, EW, McGregor,
JJ, and Murray, AM. "Data directed control and operating
systems." Computer Journal,
15, pp 125-129 1972.
Electronic Design, "Analog and
analog-digital arrays blossom." v37 n13:49-56
Epstein, Barbara. “Why
poststructuralism is a dead end for progressive thought.” Socialist Review,
pp. 83-120, 95-2, 1995.
Franks, J. Review of Gleick (1987). The Mathematical Intelligencer,
11:65-71, 1989.
Frisch, U. and Parisi,
G. "Multifractals." in Turbulence
and Predictability of Geophysical Flows and Climate Dynamics. NY: North-Holland, 1985.
Gabor, D. "Theory of Communication." J.
IEE
Gregory, Dik. "Delimiting Expert Systems." IEEE
Syst Man Cyber 16:6 pp 834-43, 1986.
Hager, S. The
Illustrated History of Break Dancing, Rap Music, & Grafitti. NY:
Haraway, Donna. Modest_Witness@Second_Millennium:
FemaleMan_Meets_OncoMouse.
Hayles, K. Chaos Bound.
Hebdige, Dick. Subculture,
the meaning of style.
Hebdige, Dick. Hiding
in the light : on images and things.
Heims, S. John von Neumann and Norbert Wiener.
MIT,
Hewitt, CE "Viewing
control structures as patterns of passing messages." Artificial Intelligence, 8, pp 323-364
1977.
Hubel, D.H. and Wiesel,
T.N. "Receptive fields, binocular
interaction and functional architecture in the cat's visual center." J
Physiol 1965, 28: 229-289.
Jameson, Fredric. Postmodernism:
or the cultural logic of late capitalism.
Kirkpatrick, SCD,
Gelatt, Jr, and Vecchi, MP.
"Optimization by Simulated Annealing." Science 220, no 4598 pp 671-80
Kolata, G. "Math genius may have hormonal basis."
Science 222:1312.
Latour, Bruno. Science
in Action.
Levy, J. "Lateral differences in the human brain
in cognition and behavioral control." in Cerebral Correlates of
Conscious Experience. ed P Bruser and A Rougeul-Buser,
Lilienfeld, R. The Rise of Systems Theory. NY: John Wiley 1978.
Lyotard, JF. The Postmodern Condition: a report on
knowledge.
MacKay, D. "Strife over visual cortical function." Nature 289 pp 117-118.
Mandelbrot BB The
Fractal Geometry of Nature. W.H. Freeman and Company, NY 1987.
Maren, A.J. "Multilayer Cooperative/Competitive
Networks," in Maren et al 1990.
Maren, A.J., Harston
C.T., and Pap R.M. Handbook of Neural Computing. NY: Academic Press
1990.
Martin, Emily. Flexible
Bodies.
McLuhan,
McRobbie, Angela. Postmodernism
and popular culture. London ; New York : Routledge, 1994
Mercer, Kobena. "Black Hair/Style Politics." New Formations #5 1988 pp 33-54.
Miller, G. "The Magical Number seven, plus or minus
two." Psychological Review
63, 81-97, 1956.
Miller, J.G. "Potential applications of a general
theory of living systems to the study of military command and control." Decision
Information, ed CP Tsokos and RM Thrall, Academic Press 1979.
Minsky, M. "The
View from 1988" - epilogue from Perceptrons, expanded edition. Cambridge: MIT Press, 1988.
Minsky M and Papert S
(1969): Perceptrons: Computational Geometry. MIT Press, Cambridge.
North, RL. "Neurocomputing: Its impact on the future of defense
systems." Defense Computing
Jan-Feb, pp 18-66, 1988.
Omohundro, S. "Efficient Algorithms with Neural
Network Behavior." report no UIUCDCS-R-87-1331, Dept computer sci., Univ.
of Ill. at Urbana.
Pollack, J. "No harm intended." Journal of Mathematical Psychology 33,
358-365, 1989.
Poster, M. The Mode
of Information. Oxford: Polity Press 1990.
Pylyshyn, Z.W.
"Complexity and the Study of Artificial Intelligence." in Mind
Design, ed J Haugeland. Cambridge:
MIT Press 1981.
Richey, SW. "The Philosophical Basis of the AirLand
Battle," Military Review, May 1984.
Rubel, L.A. "Digital simulation of analog
computation and church's thesis." J
Symbolic Logic, v 34 n3, pp 1011-1017, sept 1989.
Rubinoff, M. "Analogue and Digital Computers -- A
Comparison." Proceedings of the IRE, Oct 1953, pg 1254-1262.
Rucker, R. CA Lab. Sausalito: Autodesk Inc. 1989.
Rose, T. Black Noise. Hanover: Wesleyan University Press 1994.
Spretnak, C. "A green party - it can happen
here." The Nation April 21, 1984.
Star, SL. "The
Politics of Left and Right." in Women Look at Biology Looking at Women. Crossing Press, NY 1979.
Stoddard, R. "Washington focus -- future in the
making." Defense Computing
vol 3, #1, pp. 15, Jan 1990.
Tate, Greg. Flyboy in
the Buttermilk: essays on contemporary America. New York:Simon & Schuster
1992.
Taylor, P.J. "Technocratic Optimism, H.T. Odum, and
the Partial Transformation of Ecological Metaphor after World War
II." Journal of the History of
Biology, vol 21 no. 2, pp. 213-244, Summer 1988.
Toffler, A. The Third
Wave. NY: Morrow, 1980.
Toop, David. The Rap
Attack: African Jive to New York Hip-Hop. Boston:South End Press 1984.
Touretzky, D.S. "BoltzCONS: Reconciling connectionism
with the recursive nature of stacks and trees." Proc. 8th annual conf
cognitive science society, pp 522-530 1986.
Turkle, Sherry. Life
on the Screen. Simon & Schuster 1995.
Varela, F. "Laying Down a Path to Walking" in Gaia:
a way of knowing ed WI Thompson, Inner Traditions 1987.
Vergis, A., Steiglitz,
K., and Dickinson, B. "The
complexity of analog computation."
Technical report No. 337, Dept of Electrical Engineering and Computer
Science, Princeton University, Feb 1985.
von Foerster, H. (ed) Transactions
of the Seventh Conference on Cybernetics.
Josiah Macy Foundation 1951.
Zorpette, Glenn.
"Fractals: not just another pretty picture." IEEE Spectrum Oct
1988.
Endnotes