See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/329927931

The Practice of Light: A Genealogy of Visual Technologies from Prints to Pixels

Book · January 2014

DOI: 10.7551/mitpress/9138.001.0001

CITATIONS READS

53 210

1 author:

Sean Cubitt
University of Melbourne
184 PUBLICATIONS   688 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

1. 3 geomedia (seismometry, river monitoring, metals futures visualisations). 2. time and the nature of the image, still, moving and graphical; 3. having a short but
winderful holiday in Western Australia's south coast View project

Aesthetic Politics View project

All content following this page was uploaded by Sean Cubitt on 02 July 2020.

The user has requested enhancement of the downloaded file.

 The Practice of Light
A Genealogy of Visual Technologies

Sean Cubitt

Preface
The rewriting of the past is dynamic, oriented towards
the future. Its role is to endow the present with meaning
by offering a focus of desire to a community with
reason to doubt its future' (Debray 2004: 29)

How do visual media work? How did they get to work the way they do? Does how they work
matter? The Practice of Light ponders these questions by concentrating on the inferences of the
word 'work'. The title refers to the practice of light. Working with light, the work of light, making
things with and about light, is practice. When we speak of someone as 'practical', we think of a one-
to-one relationship with materials and tools. That privileged close relation is still widespread in
media, but the tools have become increasingly complex, and the forms of practice increasingly
involve complex interactions not only of human beings but technologies doing things, working,
making. All practice involves us in the laws of nature and, as we shall see, in learning from natural
processes, and intervening in them. The book unpacks a story of natural, human and technical
practices involved in both craft and industrial methods of accounting for or mobilising light in
visual media.

Light is the condition of all vision. The visual media are our most important explorations of this
condition. They reveal the long history of humanity's struggle to control light. The Practice of Light
presents a genealogy of the commanding visual media of the 21st century, digital video, film and
photography, tracing the evolution of their forms through a history of materials and practices.
Because of this focus, it omits the non-Western history of visual technologies, the dyes, inks,
printing and paper technologies that preceded the European trajectory of printmaking, and the
swadeshi and other indigenous forms of mechanical and digital media developed in the modern
period. Instead it traces the roots of those technologies that have become global in the 21st century.
Addressing the colonialist implications of these technologies, as well as the other face of

The Practice of Light 15.10.13 page 1

globalisation, the ecological implications of dominant media forms, will have to wait for a
companion volume to this book. It seemed important first to establish the aesthetic of dominance in
the dominant aesthetic. In the imperial era that parallels the period tracked here since the 15th
century, techniques and technologies stolen from colonised cultures have been assimilated into that
dominant aesthetic. It will take another book, or more than one, to trace the braided environmental
and decolonial histories of visual technologies. That work will look at the extent of dominance: this
investigates its intensity.

The Practice of Light begins in the invisibility of black, then builds from line to surface to volume
and space. It traces increasing degrees of complexity, passing from the simplest of marks, the line,
to two qualities of surfaces, their texture or grain and their colour. The construction of space in
visual media is addressed in the following chapter under the rubrics of shadows, layers and
projection, with perspective considered a special case of a more generalised practice of projecting
that includes, among other things, cartographic organisations of space as well as cinematic
projection. The last chapter deals with time, the most recent addition to visual culture, although in
many respects the media we can presume to be the oldest – poetry, song and dance – were always
intrinsically temporal. Time, as the movement of becoming, completes a trajectory from nothing,
the invisible dark, via something, surfaces and spaces that are clearly seen, to end in what cannot be
seen as and in itself, but yet is everywhere apparent: the time of decay and of emergence.

By media I understand the physical processes – matter, energy dimension and form – in which all
human communication takes place, including money, sex, transport, weapons and the panoply of
communication channels through which we put into practice what social scientists abstract under
the names 'politics' and 'economics'. Before we can communicate, we mediate. Merleau-Ponty
(1968) insists that the condition that he calls 'the visible' underpins seeing and being seen. Visibility
is a special condition of mediation, a term that describes the condition of everything that mediates
and is mediated in turn. Mediation is the ground of relationship, the relationship that precedes and
constructs subjects and objects. Media matter, both in the sense of giving material specificity to our
descriptions of such abstract concepts as society and environment, and in the sense of the active
verb: mediation comes into being as matter, its mattering constitutes the knowable, experienceable
world, making possible all sensing and being sensed, knowing and being known.

Light is such a mediation, not only between people, but between human and non-human worlds.
Light fills and forms the world. For millennia light from the sun and the celestial bodies, and from

The Practice of Light 15.10.13 page 2

burning organic matter derived from sunlight, spilled with the seemingly boundless generosity of a
god or gods into the human universe. It bathed our skins and filled our eyes, nurtured our food and
habitats, made it possible to see and be bewitched by beauty. There would seem to be no need to
hoard it. But everywhere is evidence that mediation is parcelled out, amassed, ossified, delayed,
hypostasised. If all human society is premised on mediation, then it is the task of media theory to
distinguish the philosophical concept of mediation as primal flow connecting all things from the
history of mediation as the engine of disconnection and delay, and thence of exploitation and
oppression. Like any other work of media history, this one must confront the dialectic of actual and
potential media: the physical media we have, and the potential that lies curled up in them.

A second principle guides this study: that among the human instincts of hunger, sex and self-
preservation, there exists an instinctual desire for order. We do not foul our own nests; we tidy our
corner of the world for habitation. This is not an entirely new thought. Sigmund Freud, in his late
writings (1961a, 1961b), identified the death drive in the repetition compulsion he observed among
shell-shocked soldiers, describing it as a biological urge towards dissolution and decay: as we
would say, to entropy. At its opposite pole Freud placed the instinct for life, attributing to
civilisation a negentropic construction of order in increasing levels, from the couple to the
household, the community and the state. This negentropic instinct for order has at one pole entropy,
but at another the excess of order, as anorexia and compulsive eating hedge round the instinctual
pleasures of food.

This principle underlies the concept of practice as it is deployed in this book. Practice involves
ordering materials and techniques. The effervescent superfluity of light is one of the entropic
instances to which we seek to bring order: our infantile, oceanic immersion in the dazzling flash of
moonbeams over rippled water, the flicker of dappled sunlight under trees, the night speckled with
stars, the fall of firelight over skin. Control over light, and its mediations through visual
technologies, matters because it alters the constitutive grounds of sensing, knowing and relating to
one another and to the world. The genealogy of visual technologies traces a historical dialectic
between the urge to control, even to fascistic excess, and the constant re-emergence of entropy in
the interstices of devices designed to curtail and command the excess of light.

Mediation is not necessarily efficient, in the sense of translating data from A to B undamaged.
Efficiency, the mantra not only of engineering but of managerialism, can serve both the instinct to
order and its excessive and obsessive application, taking on forms of regimentation worthy the

The Practice of Light 15.10.13 page 3

name of repetition compulsion. At the mass scale of relations, connections and interactions where
mediation operates, mutation is the only constant, resulting in an evolution of which it remains to be
seen whether it constitutes progress in any definable sense. The rule of order tends towards
equilibrium, but equilibrium, absolute symmetry in every direction, is also the final state of entropy,
the tendency of energy to dissipate evenly throughout space. In that final state, energy and matter
lose their shape, their clustering into higher organisations and more complex relations. As Mary-
Anne Doane (2002: 117) reports, 'the Second Law of Thermodynamics dictates what is effectively
an increase in an absence – as entropy increases, there will be less and less difference . . . in a
thermodynamic logic what is increasing is lack, or loss'. Entropy is the opposite of information, of
form. So the attempt to control mediation to an extreme results in a decrease of form, and an
increase of the very entropy it emerged in order to conquer. The efficiency gains of standardisation
addressed in the latter part of The Practice of Light also risk increasing the level of noise in the
system, while discouraging that radical change which is the hallmark of emergence. The result is a
culture that constrains us to innovation within parameters already historically established, and steers
us away from inventing the disturbing and exciting new (Stiegler 1998: 34-7). For much of the
history traced here, political economy plays a relatively small part in the genealogy of visual
technologies, but as the 19th century's development of mass media accelerated, there began a
period, in which we are still living, of increasing political-economic interest in and ultimately
control over the development of new media forms and formations. This developing role of norm
and standard as instruments of control distinguishes the modern from the pre-modern history of
visual media.

This is especially true of transmission media. In some respects, storage media had a far longer
history of separating especially prized moments of mediation – artefacts, messages, tools and
techniques – through priestly castes charged with secreting knowledge in the temple, through guild
monopolies over the mysteries of their trades, or through the amassing of beautiful things in the
palaces of the rich and mighty. In the great democratising of access to knowledge and art beginning
in the 19th century's new transmission media (chromolithography, photography and cinema among
them), celebrated by Benjamin in the 1930s, and reaching an apogee, even saturation, in 21st
century network media, those old hierarchies flattened, and their hidden treasures distributed to all
and sundry. At the same time, however, there has been a transfer of the function of control from
storage media to transmission media. We are all aware of the increasing speed of dissemination,
now approaching the speed of light, or at least of light traveling through fibre-optics, a speed which
gives us the illusion of immediacy. That speed would seem to guarantee that, eventually, all

The Practice of Light 15.10.13 page 4

mediations circulate, all secrets are revealed: information, as we say, wants to be free. Yet
information is everywhere in chains. The purpose of control over distribution is to delay
transmission. We think we pay more for premium service delivery of news and entertainments: in
fact the money pays for timely arrival, and its absence assures a deliberately delayed and often
downgraded delivery. The apparent ubiquity and velocity of transmission media hides a new class
system operating inside the new media democracy.

The crucial mechanised media of the 19th century – telegraph, photograph, phonograph,
cinematograph – share in their suffix the Greek word that stands for both writing and drawing.
Since McLuhan, media history has given the printed word pride of place, but the significance of
drawing and its early modern mechanisation in prints have attracted less attention. The history of
printmaking techniques gets a privileged place in the account that follows because it is critical to the
story of the increasingly democratic distribution of images, to the new cartographic culture of the
European empires from the 17th to the 20th centuries, and to scientific and technological
visualisations. It is in printing too that we first see the division of images into fields of dots, as well
as the introduction of scanning at the beginning of the electronic era. The persistent belief that
Gutenberg apprenticed to the Master of the Playing Cards (see Chapter 1) points us towards the
coincidence of two great inventions of the early modern period, the printing press and intaglio
printing. Many of the Master's cards are composed from smaller engraved plates, which must have
been held in a frame to compose the 'pips' of the suits. The same kind of frame held the repeated
letters of Gutenberg's printing press. Alongside the printed word, we should understand the
invention of etching as equally formative of modernity, encouraging the move from Gothic linearity
and reliance on types (the typical form as opposed to the specific individual instance) in woodcut to
the fluid, observational style of silverpoint drawing and the engraver's burin. Woodblocks allowed
longer print-runs than the shallow-cut metal plates of early intaglio printing, but had nothing like
etching's resources for fine, gestural lines and delicate shading. Though more expensive than
woodcuts, prints from metal were far cheaper and more reproducible than oil paintings, and became
the privileged medium of visual culture throughout Europe until the mass reproduction of
photographs became widespread in the later 19th century. Yet while their technological advances
assured wider transmission of images, the techniques of printmakers standardised increasingly
around schools and styles of drawing and incision even before the arrival of mass print
technologies. The demands of the new magazine and news industries of the 19th century required
further efficiencies and further standardisation in the transmission of images. One driver for this
was the mass audience which, having learned how to read the increasing numbers of images now

The Practice of Light 15.10.13 page 5

circulating, was served by an increasingly standardised lexicon of motifs, compositions and styles
for communicating the visible, such as the black line of ink denoting a ray of light addressed in
Chapter 2.

Beyond such basic visual literacy, most of us have little interest in the technologies mediating our
cultural and working lives. A major reason for undertaking this research was the realisation that I
knew next to nothing about the workings of the analogue and electronic media I had been teaching
and writing about for a decade or more. Discovering how they work is a fascinating study in its own
right. Tracing their histories gives us, in addition, ideas not only about how they work but why they
work in the precise ways they do. Observing the 'why' gives privileged access to the particular
capabilities of specific technologies, what they do best, what they do poorly, and what they cannot
do at all. Not only does this promote a move from connections to comparisons between media; it
opens up understanding of how and why particular works operate within the affordances of a
specific medium, or push at their boundaries, provide practical criticisms of their shortcomings, or
turn them into unexpected vehicles for inventions and creations for which they were never
designed.

At the same time on the macro scale, as every scholar of globalisation knows, planetary commerce
in every sense of that word depends on media. With honourable exceptions (among them Curry
1998, Fuller 2001, Tufte 2003), media scholars have been slow to address the aesthetics of the
workplace media that underpin globalisation. Picturing technologies are a relatively small part of
these political-economic media. Far more important are the databases, spreadsheets and
geographical information systems required by every company and government in the 21st century.
In parallel with workplace media are scientific instruments, increasingly tied into the networks of
power and money through earth-observation satellites (Parks 2005), biometric monitoring (Lyon
2009) and medical imaging (Cartwright 1995) among many others. These media share in the history
of visual technologies because they are themselves importantly visual, and because they share the
same architectures of storage and transmission. Workplace and scientific media are especially
important because they make apparent the increasing convergence of representational media with
media of data visualisation. Historically much of visual culture has concerned itself with making the
invisible visible, whether depicting celestial events, the lost past, the interiors of animals, cutaways
of underground structures, or infra-red and other wavelengths unavailable to unaided human vision.
The universe of numbers generated by an informational society draws on this history of visualising
to make itself visible, and in turn the particular organisational modes typical of informatics begin to

The Practice of Light 15.10.13 page 6

impinge on modes of visualisation.

A near-universal characteristic of data visualisation is its bent towards generalisation. Graphs track
both individual occurrences and the trends they point towards, averaging the contingent results of a
day's trading or a lab experiment. As a practice of generalisation, graphs have roots in both the
mathematics of abstraction and the historical legacy of types in visual culture, though the latter is
changed from the type as divine substance or Platonic idea to the type as statistical norm. Data
visualisation combines this statistical normalisation with the process of counting individual, discrete
occurrences. This textural quality, this grain of the graph, has unexpected congruences with the
chemistry of analogue photography and cinematography, explored in Chapter 3.1, where molecules
of photosensitive chemicals select photons from the flood of light and generalise them as elements
of an image. Similarly, the curve of a graph derives from plotted points defined by their coordinates
on the grid of the graph paper. To fit into the coordinate system, points must be defined by finite
numbers: units. Each pair of coordinates marks a unique spot on the graph, but every cell on the
grid is the same as every other.

This leads us to a second congruence, with film in particular but also with photography. Drawing on
Marx's contention that 'commodity is exchanged for commodity, except that this exchange is a
mediated one' (Marx 1973: 132), Jonathan Beller argues that the circulation of cinematic images is a
continuation of the same process:
the mediation between “objects” by which a film constitutes itself is
between abstractions from objects . . . the mediation between the interval of
image-commodity and image-commodity is film. It is the film frame
(screen) that allows the images to circulate: film is the money of cinema
(and the frame is the unit). The affective dimensions of capital circulation
are distilled and experienced in their most purified form in the cinema – and
in the enchanted registration of everything by money (Beller 2006: 58)
The twin processes operating here are common to cinema and to the graph as the basic form of data
visualisation. First, the object is abstracted, as image, commodity or numerical value; and second, it
is placed into relations of equivalence with all the other abstractions by dint of the formal equality
of every cell, frame or price with every other. In photography, every photograph is exchangeable for
every other; in graphing, every cell for any other; and in digital imaging, it is no longer images that
are exchangeable but their component elements, the pixels. Important to our argument is that this
abstraction and exchange belongs to the same class as the abstraction and exchange enacted through

The Practice of Light 15.10.13 page 7

money: both are mediations, and both arrive at their hegemonic rôles as a result of historical
processes. Both promise to govern time, through the measured frame-rate of projection and the
preservation of debt as control over the future.

Enumeration is a pledge against disorder. The counting numbers give us the assurance (as they
embody the metaphor) of one-to-one correspondence, whether of images in circulation or of
moneys to be repaid. But by the end of the 19th century, that old assurance was no longer
mathematically tenable, so that at the very moment the engineers were hardwiring unit-counting
into the technical media of the 20th century, theoretical mathematics was beginning to unpick it.
Kittler's precise date is disputable, but his sentiment is correct: 'innovations in the technology of
information are what produced the specificity of the discourse network of 1900, separating it from
transcendental knowledge'. As evidence Kittler offers the typewriter which 'provides writing as a
selection from the finite and arranged stock of its keyboard . . . In contrast to the flow of
handwriting, we now have discrete elements separated by spaces' (Kittler 1999: 16). Thus the
principle of the reproducible pips of playing cards passes from the printing press to the typewriter,
and in Kittler's argument becomes the basis of all subsequent discrete (as opposed to continuous)
media. According to Kittler, discrete media restructure knowledge, so that we no longer found
knowledge on the permanence of transcendental Kantian a priori categories like time and space.
Here Kittler echoes Foucault's opening attack on historical concepts of tradition, influence,
development, evolution, and spirit (esprit) in The Archaeology of Knowledge: 'instead of according
them unqualified, spontaneous value, we must accept, in the name of methodological rigour, that, in
the first instance, they concern only a population of dispersed events' (Foucault 1972: 22).

If, in Doane's version, entropy decreases difference, the negentropic order of discrete media
involves the proliferation of difference. But endlessly proliferating difference is as noisy as endless
symmetry. Mediation as the very fabric of change, of mutation, is a builder of differences, but as
bearer of communication, it also establishes organisational forms with varying degrees of longevity.
The grid of data visualisation and its forebears in the discrete mechanical media of the late 19th
century form a bridgehead to a new moment of modernity. Equally, the long migration of control
from storage to transmission, from the political economy of secrecy to the political economy of
distribution, from secreting away to placing in circulation (and charging rent on every transaction)
does not occur as a smooth transition: 'Unlike the history to which it put an end, the media age
proceeds in jerks, just like Turing's paper strip' (Kittler 1999: 18). Historiography itself falls under
the sway of the media in which it is now composed.

The Practice of Light 15.10.13 page 8

Extending Kittler, Huhtamo and Parikka (2011) hypothesise an embattled relation between their
project of media archaeology and Geistesgeschichte, the tradition in historiography that sees a
civilisational coherence (Geist, spirit) shaping all the activities of an historical epoch. But the lure
of creating order out of the archive of events is too strong, even for heroic antecedents like
Benjamin's Arcades Project (Benjamin 1999a) or Aby Warburg's Mnemosyne Atlas, whose attempt
Agamben describes as 'an “iconology of the interval”, a study of the Zwischenraum [space in
between] in which the incessant symbolic work of social memory is carried out' (Agamben 1999
100). Raiding Derrida's concept of the archi-trace in place of Foucault's archaeology, Agamben
draws out the implications of this formative difference, not only in the history of symbols but as
ontology. Citing the mediaeval divine Duns Scotus, he argues 'nothing keeps a thing from being
actual and, at the same time, maintaining its potential not to be, or to be otherwise (Agamben 1999:
262). For media archaeology, mere existence is no proof of necessity: because our media operate
just so is no reason why they have to do so, or why they may not operate otherwise. The jerky
history of media that Kittler observed is then not only a symptom of discrete media, but a
description of a problem in the history those media mediate: that in the steps between discrete
moments, what Bellour (2002) calls the entre-images, lies a formative differentiation, an ontological
instability, the Zwischenräumen from which the architect builds his house in Morgenstern's poem
(cited in full, less than coincidentally, in Kittler 1990: 256-7).

At stake here is the primacy of Being, and of being-one, the root condition of discrete media.
Countable objects require two things: an object to be counted, and a subject to count. But if the
object is to be one, it must be whole. Ontologically, its existence as object requires a subject to
recognise its objectivity, a reciprocity that already undermines the unity of the object. But to be
counted, the object has to be both itself and its representation as a number, which likewise divides
it, even before the act of enumerating places it in numerical relationship with other numbers and
their objects (as a name not only adds itself to the thing it represents but ropes the named thing into
the chains of language). This primary instability of objects when they are mediated reveals the
mediation that generates them as objects in the first instance, and in the case of discrete media
posits their relationship with other objects as one of difference. In this contradictory structure of
existence lies the germ of the historiographic premise of the current volume. We cannot but write
the history of media as a history of discrete events, but in doing so we must insert gaps between
those events, gaps that, as differences, generate further instabilities. Thus history is not a plenum but
a proliferation of interruptions, each of which generates comparisons and continuities (of the kind

The Practice of Light 15.10.13 page 9

investigated by Warburg) as well as the unrepeatable isolation of events that might always have
occurred otherwise, with other consequences. As Parikka (2012: 2) writes, the spirit of media
archaeology is 'thinking the new and the old in parallel lines'. In this perspective, the continuum of
history lies not in the chain of events, but in the spaces between them, the grit and noise, the heat-
generating friction of event grinding up against event.

The question of the continuum is both a historiographical and a technical one. As Michel Serres
(2007) argues, noise is intrinsic to any system. Not just the external interruptions but the
interruptions intrinsic to any signal, noise is the permanent condition of communication. Without
the static of solar and cosmic radiation, we would have no radio. The continuum is just this
bubbling up into existence of potentialities, realised or unrealised. At the same time, of all the
mutations that occur randomly, we know that some have the greater likelihood of establishing
themselves, of gathering probability that they will recur. Media history is not a matter of the
ontological grounds of being or becoming, but of the actual events and situations that emerge from
it: objects are no less material for the fact that they have to be brought into existence, and are
perpetually interconnected with the past potentialities that generated them and the present universe
of objects with which they form their networks. Enquiring into the materiality of media, the
minutiae of their operation, exposes both the contingency of their existence, and the role of
probability in bringing this rather than that into dominant position. It is this question of dominance
emerging from the static, of the congealed forms of dominant media and their associated ways of
perceiving and knowing, that differentiates the genealogical project of The Practice of Light from
those forms of media archaeology that emphasise the paths not taken. Although The Practice of
Light does speak to at least one lost opportunity, the vector screen, its major task is to explain why
we have the media we do. Having discounted the ideas of progress and efficiency, there remains the
thesis that our media are simply the most probable of all alternatives.

The second feature of data visualisation noted above was the process of averaging. The tendency
towards enumeration belongs to the world of the commodity; averaging however belongs with a
relatively new form of government, with power, and therefore with politics. There is an attempt here
to re-introduce to one another the two halves of an integral discipline, political economy, torn apart,
as Vincent Mosco (2009: 45-64) has it, in the attempt to make a science out of economics by
ignoring the messy business of political life, and further severed in the understandable but
unfortunate anti-economistic stance of Foucault. For Foucault, biopolitics concerns population,
a multiple body, a body with so many heads that, while not infinite in number, they cannot be

The Practice of Light 15.10.13 page 10

definitively counted. Biopolitics deals with the population as a problem that is at once scientific and
political, as a biological problem and as power's problem (Foucault 2003: 245) Unlike earlier
formations of sovereign and disciplinary power, biopower works on series of probable events,
determining 'an average considered as optimal on the one hand, and, on the other, a bandwidth of
the acceptable that must not be exceeded' (Foucault 2007: 6). Speaking of biopolitical town
planning, Foucault emphasises key qualities of the new form of rule:
It is simply a matter of maximizing the positive elements, for which one
provides the best possible circulation, and of minimizing what is risky and
inconvenient, like theft and disease, while knowing that they will never be
completely suppressed. One will therefore work not only on natural givens,
but also on quantities that can be relatively, but never wholly reduced, and,
since they can never be nullified, one works on probabilities (Foucault 2007:
19).
To achieve the enumeration required by both commodity exchange and digital media requires the
reduction of the messy, noisy world to units by a process of stabilisation through statistical sampling
and probabilistic analysis. Among the techniques analysed in The Practice of Light, the capture of
light, transmission of signals, and the compression and decompression of video provide powerful
examples of this process. Foucault himself emphasises only the media of language and archives:
Kittler expanded his terms to technical media and their operation. Here I argue that these principles
of statistical probability and unit enumeration govern not only power and money but the minute and
constant operation of visual media. It is at this level, as much as through Foucault's argument about
the mutual implication of power and economy in liberalism and neoliberalism (Foucault 2008: 131,
202), that The Practice of Light propounds the reconstitution of political economy at the
foundations of visual aesthetics. Aesthetics, the sensations through which we sense the world, are
shaped by the technologies through which we do that sensing. True, we are not dupes; and true,
technologies do not spring fully formed from the mind of God. Quite the contrary: the millennia of
struggles over how to order mediation into communication shape and constrain those
communicative technologies which in turn shape and constrain our sense of the world. Aesthetics
concerns the struggle for control over the human sensorium, and thus over the ordering of
communication which is the order of our polities, societies, cultures and economies.

One important example is the resolution of the long-running dispute over physical and
psychological understanding of colour described in Chapter 3.2. The establishment of a normal
observer has been critical to the mass manufacture of visual technologies. In the early decades of

The Practice of Light 15.10.13 page 11

broadcasting, radio and TV signals competed for airspace with military, transport, government and
emergency service users. The large tranches of bandwidth required by early transmission
technologies made bandwidth a scarce resource, and many technologies developed techniques to
make do with the restricted quality (and unreduced noise) of their systems. To do this, they
developed not only engineering solutions, of which the most significant is enshrined in Shannon
and Weaver's (1949) mathematical theory of communication, but psychological models. Studies of
hearing discovered that listeners could make out the words of a conversation even in a noisy
environment: therefore telephone suppliers could afford to produce only low-quality signals, on the
basis that a statistically normal user would be able to get enough information down a noisy line to
understand the message. Applied to visual technologies, this principle allows transmitters and
networks to employ the lowest level of clarity and fidelity needed for an end-user to make out the
form and colours of an image. The good-enough principle rests again on the foundations of
statistical norms first formulated in Quetelet's 'social physics' of l'homme moyen in the 19th century
(Desrosières 1998; Hacking 1990). Far from leading towards the best possible resolution, the
evolutions of media technologies analysed in the following chapters commonly standardise on the
lowest acceptable solution, where 'acceptable' means conforming to a normative psychology of
perception, and a firm belief in the privilege of message not only over noise but over the inessential
qualities, contextual, poetic and emotive, that would otherwise colour and shape our sense of
communion. In the interests of efficient management, and employing the tools of probability and
norm, our media are also practices in the sense of Iago's “practicing upon his peace and quiet”
(Othello II, i, 235), practicing on our perceptual gullibility, while shaping our expectations around
the efficient-enough delivery of data.

In the statistical management of sensory data and its unit enumeration there emerges a diagram of a
state of affairs which, while actual in its instrumental application in visual technologies, contains
both disturbing and liberating virtual potentials. Adorno points out,
Exchange is the rational form of mythical ever-sameness. In the like-for-like
of every act of exchange, the one act revokes the other; the balance of all
accounts is null. If the exchange was just, then nothing should really have
happened, and everything stays the same. At the same time . . . the societally
more powerful contracting party receives more than the other. By means of
this injustice something new occurs in the exchange: the process, which
proclaims its own stasis, becomes dynamic (Adorno 1998: 159)
going on to note 'the ever-sameness of the exchange principle intensifies by virtue of technology

The Practice of Light 15.10.13 page 12

into the domination by repetition in the sphere of production' (160), a process which in this work is
traced further into the reproductive mechanisms of visual culture, from print to frame to pixel. Yet
Adorno is also keen to demonstrate that this injustice that drives progress (which we might translate
in the language of neo-liberalism as 'growth') is also the grounds for its eventual disappearance:
constantly completing mendacious contracts based on broken promises of equal exchange is
ultimately indistinguishable from the abolition of exchange. In this sense, the reunification of
politics and economics in the realm of aesthetics points towards more than a better, if bitterer,
understanding of the nature of our oppression. Though not utopian, this book's analysis drives
towards a politics of hope.

The problem of exchange as model extends into the mutual mediation of subjects and objects,
ossified in the commodity form as a separation without mediation in which 'Once radically
separated from the object, subject reduces the object to itself'. We should not imagine a primal
blissful unity before this separation, however: 'Before the subject constituted itself,
undifferentiatedness was the terror of the blind nexus of nature' (Adorno 1998: 246). The
constitution of the subject as individual is repeated in every individual, the myth of individuality
undercut by its compulsory severance from its objects and the resulting compulsory engagement in
economic life (Foucault 2004: 202) and in communication (Dean 2005), a tendency marked
lexically in the move from the older '-graph' suffix to now our common usage of 'audio' and 'video',
literally 'I hear', 'I see'. Here we must part company with the philosophy of flux, the Bergsonian
accusation that cinema breaks up the continuity of duration, and accept the tragic rift between self
and world, populations and their environments, mediated through the necessary cut that redeems us
from the formless terror of absolute indifference. At the same time however, we have to understand
that contemporary media replace the chaos of indifferentiation with repetition, symmetry, and the
homeostasis of endless exchanges summing at equilibrium. The vast humming of the technological
noosphere is the perverse utopia of a market that would be both just and equal.

Ordering has become both separation and repetition, governed by probabilities that tend towards the
equilibrium of the least worst. If Adorno is correct, and the original sin that starts this process is the
subject-object relation, nonetheless the mediations of object and subject have a history, and we must
start to look for a third term, a project. A third quality of the graph is its temporality, at least in those
common graphs with time t as an axis. While the cells of graph-paper demarcate progression in the
calculative units of clock and calendar, they preserve at least a memory of another form of time, the
gesture that marks a curve onto a surface with charcoal, pencil or brush. This gestural continuum I

The Practice of Light 15.10.13 page 13

refer to here as the vector, a privileged form in computer graphics whose longer cinematic history I
traced in The Cinema Effect (Cubitt 2004). The vector is the dialectic of multiple and changing
forces operating on a movement through time. As dialectic, the vector also generates the forces
operating on it: the swoop and swerve of a line is the joy of becoming that generates the action of
the hand, the tool it moves, the mark itself and, as Hogarth was so prompt to observe, the motion of
the eye that follows it. The term describes not just a type of line, but a principle observable in
camera movements, fly-throughs in digital animations, and in the continuities built through the
succession of images. In this it imitates Merleau-Ponty's (1968: 264) concept of the chiasm as an
interleaving of world and subject, 'the insertion of the world between the two leaves of my body /
the insertion of my body between the 2 leaves . . . of the world' and at the same time the motion of
suture, 'the general relation of lack to the structure of which it is an element' (Miller 1977/8: 26; see
also Butte 2008), the principle of difference, the non-identical, the non-existent from which
becoming springs. The vector is no escape from rule, but the constant generation of new rules
connecting subject and object in the creation of new forms of mediation. It confronts the imposed
order of symmetry and stasis with constantly mutating, improvisatory motion into an unforeseeable
future. In the nexus of all actually existing mediations, vectors emerge as their virtual dimension,
their capacity to change into something other than the already known, the already experienced, the
existing map of relations. Where the graph defines itself as the aggregate of points, vectors are
continuous, and continuously in metamorphosis, continuously negotiating the forces, situations and
events they encounter, and continuously redefining and changing those forces, situations and events
and the relations between them which it mediates.

Technically, a vector is a mathematical description of a movement through time. The description

given here is based on the principle that that movement, that gesture, is open-ended. In some
contemporary media practices, notably in codecs, the vector is bounded by start and end points.
Even though the same maths describes its motion, this instrumental vector is no longer open. As
Bazin lamented, no technique is uniquely destined to produce one effect and one effect only. Similar
contradictions between utopian and dystopian aspects afflict almost every technique investigated in
this book. Even the unit, averaged grid which, posited as hegemonic, plays the villain's part here,
harbours internal contradictions which enable works of great wisdom and beauty, a handful of them
described in the following pages. Techniques and technologies are neither stable nor immune to the
principle of non-identity. When Beller argues that projecting film at 25 frames per second orders
perception in a particular way, he does not argue that all films are therefore the same. The TCP/IP
suite governing internet traffic does not make all websites identical. A technology is a mode of

The Practice of Light 15.10.13 page 14

mediation, a relation. In Scott McQuire's words (2006: 252), 'defining the technological activates
the border between nature and culture'. What we call power and economics can also be defined as
mediations between populations and environments, and this because, when we speak of relations of
domination, 'We do not try to trace their origins back to that which gives them their basic
legitimacy. We have to try, on the contrary, to identify the technical instruments that guarantee that
they function' (Foucault 2005: 46). Those instruments are media, taken, again, in the broadest
definition: the media of mining for the materials that build computers, as much as the media of
police weapons trained on miners, and the financial flows from mines to corporate offices and R&D
labs; the media of farming pigments and glues, as well as those regulating the clearing of rainforests
and of transport and trade.

Practice is never individual; nor is a technology. According to Deleuze 'The machine is always
social before it is technical. There is always a social machine which selects or assigns the technical
elements used' (Deleuze and Parent 1997: 126-7). The social machine is not 'society' or some other
imaginary force. It is an assemblage of actors and mediators in Latour's sense, an interconnected
web of mutually influencing and translating activities, each open to contingency, but in the
ensemble capable of acting as a single, ordered machine. The 'traceable associations' that Latour
(2005: 106) pursues form an ensemble of interactions, the material forms of working together that
makes any kind of social life possible. Thus the materiality of mediations are technically and
environmentally as well as humanly connected, in turn connecting, informing and construing what
we understand by human, technical and environmental.

Technologies are neither autonomous nor inhuman. Any technology assembles not just matter and
energy but information. It organises the skills and practical knowledge accumulated by our
ancestors into physical form. Marx's 'dead labour' is our common heritage (1973: 690 ff; Cubitt
2008a), the living presence of the innumerable dead. When Stiegler (1998) speaks of the 'originary
prostheticity' of human being, he posits that to be human is definitionally to be a user of tools, thus
to be born into an insufficient body that requires, for its survival and for its relation with the world,
the supplement of technology. This implies also that tools mediate a relation with an environment
which is inevitably altered in its meeting with the human population. Their relation is always
unstable, an instability out of which grows the history of that relation. Though the epigenetic
principle that characteristics acquired by one generation can be passed to the next may not apply to
all species, it does to humans. From earliest times our linguistic, musical and visual media have
persisted across generations, bearing their freight of acquired wisdom while also of necessity

The Practice of Light 15.10.13 page 15

passing on the material practices of those media technologies. Every technique and medium is the
result of accumulated knowledges and skills.

In writing genealogy, however, accounting for every mediation is impossible. Nor is a statistical
sampling of data compatible with the problem we have already encountered, when historiography is
shaped by the medium in which it is enacted. Two principles therefore guide the genealogical
method undertaken here. The first is to make the comparisons and connections made possible by the
persistence and maintenance of skills across generations: to relate the artefacts of one era to another,
in this imitating the processes by which technologies, as accumulations of ancestral skills and
knowledge, perpetually compare, contrast and reassemble elements derived from the great
storehouse of the past. The genealogy of hegemonic forms of mediation in the 21st century shares
with media archaeology a concern for formations sidelined or abandoned en route, such as the
vector screens discussed in Chapter 4. The specific trajectory of evolution that occurred is not
presumed to have been necessary, but genealogy has to observe that the fact that it did occur grants
to hegemonic media formations the privilege that comes with having actually existed. Nonetheless,
the potentialities cast off along the way contribute to the definition of the actualities that become
dominant, as do specific ways of working within the dominant evolution of media forms.

This leads to a second methodological principle: to work at the level of the material specificity of
individual instances of mediation. The humanities are perhaps the last redoubt of this method,
increasingly dismissed elsewhere as 'anecdotal', that insists on the unique significance of the
particular, especially when we aim to discover the singular properties of a specific practice. The
anecdotal method runs the risk of giving the false impression that individual artists, artisans and
engineers originated specific forms and techniques: this clearly is incompatible with the thesis that
all technique is the fruit of ancestral knowledge. Nonetheless, every act of making, including those
in serial form and those using automated processes, is a new concatenation of forces, and in
potential always the birth of a new vector. That is why this book does not offer a chronological
survey in the manner of the excellent work of Briggs and Burke (2002), Debray (1991), Mattelart
(2000) and Winston (1998) among other media historians. Such histories must in some sense be
representative: to recount many instances, and to sum their common traits. The events – inventions,
artworks, theories – singled out in The Practice of Light are representative only in the sense that
each stands at a particular crossroads, drawing together the threads of the past in a unique pattern,
and projecting forward through time a specific trajectory.

The Practice of Light 15.10.13 page 16

The term 'mediation' then refers both to the principle and to specific acts of mediation, much as we
speak of 'film' and 'films'. Glory presents a genealogy of mediation as the ground of human and
non-human action, the indefinite article suggesting that there are others. The term 'practice of light'
evokes many mediations, many orderings, many events that I have not been able to include:
theological or quantum theories of light, the neurophysiology of vision, pyrotechnics, stained glass,
domestic and public illumination. Those I have selected, as selection is inevitable even in broader
surveys, I chose in order to emphasise the singularity of each moment in the longer becoming of the
architectures of visual communication in my own time, and their mediation – imitation, adaptation,
critique, subversion, misunderstanding even – of earlier forms. I have tried to orchestrate these
selections and arguments on the basis of three virtues to which I aspire, consideration, wonder and
hope: to consider in as much complexity as can be achieved the precise configuration of visual
culture in particular mediations; never to abandon to habit or cynicism the ability to wonder at the
intricacies of each mediation; and to embrace, even against all the odds, the hope that, if we have
built ourselves a demeaned perceptual repertoire, even within its structured regime there are
inefficiencies, frictions, noise and contradictions that can still generate the genuinely new, and that
beyond its borders, other modes of seeing and being seen promise to emerge.

In turn the chapters address the key themes of the book: invisibility and the non-identical, geometry
and the vector, enumeration and averaging, apparent versus virtual, and the struggle to control
light's chaotic flow. Chapter 1 argues that the difficult pursuit of black as an effect as well as a
material reveals a fundamental instability in the process of making visible. From his representation
in cinema to his emulation in Picasso's printmaking practice, Rembrandt van Rijn is the tutelary
figure of the chapter, which is occupied with attempts to achieve pure black, especially in
printmaking, through techniques that involve not only the production of complex materials and their
handling, but also developing ways to trick the beholder into seeing blacks that do not actually
exist. If black is the marker of an absence – of light – then absence has also been a marker of black.
This ontological and phenomenological instability remains integral to the visualisation of black
from Gutenberg's inks to early greyscale television, and provides a first inkling of why it is that
visual technologies have to pursue forms of ordering that instability.

Chapter 2 traces the rise of geometry as a governing principle in visual technology by looking at
five instances where key transitions have been effected. Strongly represented in Rembrandt's famed

The Practice of Light 15.10.13 page 17

collection of prints, Dürer's accommodation of observation to the grammar of proportion, especially
in the observational drawings and prints of animals and plants, opens drawing to a Euclidean
geometry that would very soon inform the proto-scientific investigation of optics undertaken by
Descartes. In the plates to Descartes' Dioptriques, as in those to Dürer's Painters Manual, there is
both an increasingly rationalist account of light as linear and instantaneous rather than pervasive
and flowing, and at the same time a growing irrationalism. The terms of this contradiction come to a
brief but influential resolution in Hogarth's roccoco line, and its expression in his treatise on The
Analysis of Beauty. This sinuous, sensuous line informs, only to be disciplined in new ways by, the
quasi-industrial line of Disney's animation studios in the 1930s, but re-emerges in the form of vector
graphics, represented here in the plotter-printer works of Roman Verostko. The instability and
fluidity of the line is both blessing and challenge, one taken up very differently in the construction
of space, time and, in Chapter 3, of surface.

Chapter 3 follows the emergence of two central features of digital imaging in the industrialisation of
printmaking in the 19th century. Apparently innocuous techniques for preparing plates for printing
by stippling their surfaces gave way to other print technologies operating on the principle of a
random scatter of dots, including the most revolutionary of them all, photography. The dot for the
first time begins to take on the strategic dominance previously occupied by the line. Splitting
images into spots enabled rapid printing, but it took the invention of scanning to tame those dots
into a new kind of line produced by scanning the entire picture plane. The earliest electronic
transmission of images by fax and wire introduced the principle that governed the invention of
modern television broadcasting. But like photography, early TV screens were composed of random
fields of spots and dots. Photography was tamed by the introduction of halftone printing; television
by a series of inventions culminating in linear scanning, an architecture traceable at the molecular
and even quantum level. From cameras and to data projectors, images are structured from the
moment of their capture in a grid of cells, identical save for their numerical designation. This
structure takes us to a point where the commodity form of the image is ingrained at the finest levels
of their genesis, transmission and consumption.

The flux of light does not immediately lend itself to this arithmetic handling. It requires an
intermediate step, traced in the way optoelectronic chips average the light they receive before
converting it to whole numbers. This process of averaging, which occupies Chapter 3.1 on colour,
begins in the slow, and as yet still not quite complete abstraction of colours from their material
sources and the grammar of colour that structured early modern works like van Eyck's Arnolfini.

The Practice of Light 15.10.13 page 18

The process was dialectical: Newton's mathematisation of the rainbow clashed with Goethe's
phenomenology of perception, and the struggle between physics and psychology would not be
resolved until the application of statistical methods to perception in the 1930s. Colour printing
meanwhile would move from the floating colours of chromolithography, beloved of Walter
Benjamin, to increasingly controlled systems, culminating in the wholly abstract colour charts of
industrial suppliers in the late 20th century, and in electronic colour management in the 21st. The
integration of statistically managed perception with numerically coded grids produces the surfaces
which now populate so much of 21st century urban and intimate landscapes.

Traditional pictorial surfaces can only suggest the existence of volume through techniques like
texture and shading, the practice that opens Chapter 4. In digital imaging, volume and space can
exist independently of the surface depictions we recognise on screens. The chapter traces the
construction of space first in shadows, responsible for volume; then in layers, as a practice of
creating the illusion of receding space; and thirdly projection, the coordination of virtual and actual
space. There has been an abiding lore of attached shadow, the shadows an object casts on itself, of
an eyebrow over an eye for example, or the shadow formed on the sunless side of a tree trunk. Cast
shadow however has a patchier history, tied into the tale of geometrical organisation of space. Most
of all, shadow colour, observed as far back as the Renaissance, which has to deal with the multiple
reactions of light and surfaces in continuous space, has proved elusive, a complexity too far in many
systems. As with colour, we discover again the readiness to apply 'solutions' that simplify
representation to the point of bowdlerising it. Something similar can be said of layers, the technique
of segmenting space into flat fields, stacked to give the appearance of distance. Deriving from
theatre, the use of flats in films, and of layers in digital imaging, organises space into discrete and
discontinuous tranches. Layers realise what is implicit in the gaps in shadow handling: that space
becomes increasingly discontinuous in contemporary practice. The same should not be true of
projective space, and yet even here the struggle goes on between modes of organisation. The
geometric principle of perspective emerges from the religious view of optics and the order of the
world, but is rapidly subordinated to the disciplinary function of the Mercator projection, and to the
automation of Albertian perspective in the design of lenses. The purest moment of this projective
ordering of space is Speer's Zeppelin Field, but its most pervasive from is in fibre optics where, at
the heart of telecommunications, we find projected light rendered coherent, and binary. The
principle of segmenting space into discrete units is not the only route however. The chapter
concludes with a genealogy of vector space since the experiments of the Cubists, and with the
artefacts produced by the transformation of continuous and mutable space into raster displays.

The Practice of Light 15.10.13 page 19

 Raster screens also organise their imaging in time. Chapter 5 traces the relations between
chronophotography, cinema, television and digital video as modes of temporal order. The continuity
between still photography and film has been overstated: film's unit is the succession of frames, not
the individual frame. Arguing that the photographic still is always already unstill, it suggests that
film took as its task to overcome this peculiar property through its rapid substitution of one unstill
still with another, until it could emulate the instability of the world it observed. Broadcast
television, especially live TV, increases the incompleteness of each scanned image, sacrificing the
structuring roles of beginnings and ends for the 24/7 cycle of endless image streams. This technique
for settling the unstable, however, still generates its own excess. New compression-decompression
systems of transmission restructure the image, not only in the raster grid and colour management,
but in slicing the plane of the screen and carving the flow of time into discrete segments. This
segmentation structures time as a succession of whole moments, each discrete from one another and
from the continuum which cinematic vision had come to visualise. In the works of certain artists
operating close to the code level of technical imaging we can sense emerging an alternative
aesthetic, here named the vector, with a very different orientation to time, one that draws on
discarded elements of the tradition traced throughout this book. It is characterised in contrast to the
assertive actuality of the informatic mode of vision enshrined in contemporary digital video.
Against units and averages, it poses multiplicities and singularities; against the apotropaic solidity
of segmented space and time, it embraces the non-identical, source of the virtual capacity of world,
media and viewers to become other than they are.

The unstable dependence linking seer and seen is one engine of the history of visual technologies.
The universe projects its ancient light onto and through our sensorium, while we project our order
onto it. The history of orderings traced in these pages suggests a deliberate forgetting of these
foundations of perception. Analysing the dimensionality, the space and time construed in the
genealogy of visual technologies traces that loss in order to understand how it has come about and
how we might remediate our relationships with the perceivable universe. At one time the conclusion
to the book was to be titled 'White' but, on reflection, white no longer appeared to be the only or the
most telling contrast to the black with which it opens. Instead, understanding light and its
mediations as singularities and multiplicities gives us the grounds for an understanding of the
ethical and political dimensions of communication, not as means but as goal.

The Practice of Light 15.10.13 page 20

View publication stats