Word & Image

A Journal of Verbal/Visual Enquiry

ISSN: 0266-6286 (Print) 1943-2178 (Online) Journal homepage: http://www.tandfonline.com/loi/twim20

Diagram: Deleuze’s augmentation of a topical

notion

Kamini Vellodi

To cite this article: Kamini Vellodi (2018) Diagram: Deleuze’s augmentation of a topical notion,
Word & Image, 34:4, 299-309, DOI: 10.1080/02666286.2018.1473145

To link to this article: https://doi.org/10.1080/02666286.2018.1473145

Published online: 29 Nov 2018.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=twim20
Diagram: Deleuze’s augmentation of a topical
notion
KAMINI VELLODI

Abstract The concept of the diagram, in its common understanding as a representational schema, has enjoyed renewed attention in
recent years across the humanities and the social sciences, bolstered in part by the ‘spatial turn’. In line with this trend, recent writing
within aesthetics, art theory, and art history has enthusiastically embraced diagrammatics as a tool for analysis. These readings tend to
incorporate the diagram within formalist and representational modes of analysis. In contrast, Gilles Deleuze presents a non-
representational conception of the diagram that offers a way of thinking of the genesis of works of art as a ‘being of sensation’, the
nature of art’s thought as constructive, and the relation of art’s work to art’s history through the ontological foregrounding of
‘difference’. This article sets out these two contrasting positions on the diagrammatic, with three principal aims in mind: (1) to indicate
and diagnose the currency of the diagrammatic as a concept within aesthetics and in the analysis of artworks; (2) to place Deleuze
within, whilst distinguishing him from, recent readings of the diagram within art history and visual culture, including those of Ernst
Gombrich, W. J. T. Mitchell, Benjamin Buchloh, Margaret Iversen, and Frederik Stjernfelt, and in so doing to expose the way in
which Deleuze’s philosophy augments wider debates in the philosophy of art; and (3) to unpack and expand critically the role and
status of the diagram within Deleuze’s philosophy of art and indicate its implications for aesthetics and art history.

Keywords diagram, Gilles Deleuze, Ernst Gombrich, form, representation, schema

better.2 For him, a diagram is the means by which ‘any

The lure of the diagram
course of thought can be represented with exactitude’3, and
‘Diagram’—from the French diagramme, the Latin diagramma,
this ‘thought’ is the rational and logical mode in which all
and the Greek diagramma, meaning a ‘geometric figure, that
men are able to participate.
which is marked out by lines’ (from (dia-) ‘across’ and (gra-
The diagrammatic has been the subject of a growing intel-
phein) ‘write, mark, draw’)—is conventionally taken to desig-
lectual interest over the last few decades, one that shows no
nate a type of map or a schema that represents in simplified
sign of abating. Some, like Peirce, have seen the diagram as a
form a state of affairs. As ‘an illustrative or geometric figure
mental representation. For others, it has served as a material
which […] gives an outline or general scheme of its object, so
tool. Since the 1980s, the continental-oriented humanities and
as to exhibit the shape and relations of its various parts,’1
the social sciences have witnessed a proliferation of new dis-
functioning as a tool for grasping information more efficiently,
courses on, and practices of, the diagrammatic, with figures as
as a means of reasoning, or as a pedagogical device, the
apparently far- ranging in their work as the architect Peter
diagram and diagrammatic thinking would appear to be utterly
Eisenman and the philosopher-mathematician Gilles Châtelet
entwined in our day-to-day lives.
linked by their preoccupation with the concept.4 The so-called
The diagrammatic also as a rich intellectual history—not
‘spatial turn’, the post-Heideggerian shift towards the notion of
only as a subject but also as a method, and not only in the
space as a descriptive, analytical, and critical tool, was no
history of mathematics, where, from Euclid to John Venn, it
doubt at the source of this interest. This move was in part
has played an obvious role as a geometrical figure, but also
fuelled by the critique of nineteenth-century historicism in the
across the broad sweep of the history of ideas. From Plato’s
wake of the apparent accentuation of the social, political and
use of a diagram to teach the Pythagorean Theorem to
cultural ‘significance of location’. It was further bolstered by a
Nicholas de Cusa’s metaphor experiments, from Villard de
series of twentieth-century intellectual waves—the explosion of
Honnecourt’s figures to the schematic apparatus of Filippo
logic in analytical philosophy and its absorption into the con-
Brunelleschi’s linear perspective, from Alexander Kojeve’s
tinental humanities, twentieth-century phenomenology’s fore-
diagrams of the history of philosophy in his celebrated lec-
grounding of visual experience and sensation, the figural
tures on Hegel to Henri Bergson’s diagram of a cone, from
thinking of Gestalt theory, the interest in diagrams within semio-
Jacques Lacan’s famous ‘knots’ to Lewis Carroll’s squares, we
tic theory (such as Algirdas Greimas’s structuralist square, and
do not need to look far to see the ubiquity of this fertile
Peirce’s labyrinthine architectonic), the problematization of
notion. The nineteenth-century American philosopher
writing and the graphic sign in the wake of Jacques Derrida’s
Charles Sanders Peirce described the appeal well, defining
work (a field of inquiry within which the concept of the dia-
the diagram as a sign which, through the suppression of ‘a
gram, with its suggestive sense of ‘across-writing’ plays an
great quantity of details’, allows the mind ‘more easily to
intriguing role), and the theorization of word/image relations
think of the important features’ and therefore to reason

WORD & IMAGE, VOL. 34, NO. 4, 2018 299

https://doi.org/10.1080/02666286.2018.1473145

# 2018 Taylor & Francis

within the study of cultural representations (Nelson Goodman, fundamental datum, privileged reality or homogenous medium of
Roland Barthes, W. J. T. Mitchell, Ernst Gombrich)—to name expression, it would seem that Deleuze’s diagrammatics does not sit
but a few. comfortably within the ‘spatial turn’. It is instead best understood in
Such intellectual shifts, given renewed significance by the terms of his philosophy of a ‘transcendental empiricism’, a philoso-
‘spatial turn’, have projected the diagrammatic, and its allied phy that attends to difference as the regime of heterogeneous, pure,
notions such as cartography, mapping, and even ‘picture’, into constructive, spatio-temporal singularities.10 It is in this light that I
the foreground of the intellectual arena.5 In the current age of propose to read Deleuze’s diagram and assess its untimely implica-
global networks, big data, and the technological circulation and tions at a remove from contemporary theoretical preoccupations
proliferation of images, an age in which the fact of the hege- and in its potential for the analysis of artworks.
mony of the visual seems indubitable, the primacy of the spatial
as a means of making sense of human experience has retained Diagrammatic art as schematic representation:
its potency. It is no surprise, then, that diagrammatics, a con- Gombrich, Peirce, Kant
cept that traverses the major intellectual fields of the huma- Many would in fact argue that the recent ‘diagrammatic turn’
nities, including semiotics, logic, linguistics, phenomenology, is really no shift at all, since the appeal to diagrams to aid
metaphysics, and aesthetics, continues to exert its fascination. thinking is as ancient as man’s awareness of his powers of
Predictably, given its apparent subscription to ‘spatial’ con- reasoning. This appeal also characterizes the process of artistic
cepts that include, together with the diagrammatic, the notions production. Artists, too, think diagrammatically, and perhaps
of ‘cartography’ and ‘geo-philosophy’, the philosophy of Gilles have always done so.
Deleuze and Félix Guattari has also been associated with this This is the chief argument made by Ernst Gombrich in his
spatial turn.6 Over the last few decades, Deleuze’s concept of seminal Art and Illusion (1960). Like Peirce, Gombrich felt that the
the diagram too has been the subject of growing interest.7 appeal to diagrammatic formulas, or ‘schemas’ (he uses the
However, this apparent timeliness is all the more reason for terms interchangeably), was a ‘common human trait’ that indi-
emphasizing its untimely character, the way it—like all effective cated ‘the tendency of our minds to clarify and register an
philosophical and critical concepts, according to Deleuze— experience in terms of the known’.11 For Gombrich this trait is
resists identification with the present age as a tool for commen- given compelling expression in the artistic production of pic-
tary or reflection upon it.8 Indeed, Deleuze’s position must be tures. As such, pictures—which are for Gombrich not just men-
distinguished from the appeal to the diagrammatic by recent tal diagrams, but physical realities—reveal something
theorists, such as Mitchell, Edward Soja, and John Pickles, fundamental about human processes of thinking and perception.
where we find it being used to describe aspects of the ‘con- Art and Illusion remains one of the most cogent defences of
temporary’ lived state of affairs—whether this be the hege- the claim that pictures in the Western tradition are represen-
mony of images and the ‘pictorial turn’ (Mitchell) or the tational. And it is Gombrich’s nuancing of the question of
spatial ‘transformations of the contemporary world’ through representation as a schematic, or diagrammatic, act of artistic
the flows of capital and goods, global immigration, tourism, construction is particularly notable. Pictorial representation,
and cyberspace and the internet (Soja, Castells, and Pickles). he argues, is bound with the processes of representation
Such positions invest the diagrammatic as an agency for the inherent in man’s perception, where perception is defined as
representation of a contemporary lived situation, a means of the registration of sense experience human perception.
rendering intelligible a complex state of affairs. But Deleuze’s Pictures are representational, that is, not in terms of their
philosophy of the diagram departs from such a logic of intellig- illusory fidelity, or imitative approximation, to nature, and
ibility and representation.9 For him, the crucial thing about not in terms of the way things are directly ‘seen’. Rather, they
diagrams is their transformational character. To map a situa- are representational with respect to their appeal to schemas,
tion is not to show it more clearly, but to change it. ‘the basic geometric relationships which the artist must know
As such, the very notion of ‘space’ that is implicated by diagram- for the construction of a plausible figure’.12 On the basis of
matics alters in Deleuze’s account. If to map a situation is to bring a this theory Gombrich sets out the history of art as a history of
new situation into being, then space cannot simply be considered a schematism, narrating a fascinating chronology of the dia-
homogeneous medium for the circulation of heterogeneous objects grammatic that takes in the pattern books, plans, scribbles,
(pictures, images, flows, data, goods etc.). A diagram is not a and sketches of artists including Villard de Honnecourt,
mapping of things that can already be compared. Instead, Albrecht Dürer, Leonardo da Vinci, Peter Paul Rubens,
Deleuze speaks of distribution of singularities before the emergence Alexander Cozens, and John Constable. Through these case
of the spatial and temporal coordinates that define the successive studies, Gombrich demonstrates how artistic perception and
positions and reference points of an object. That is, diagrams are the construction of style have always been a matter of regis-
genetic: they bring new reality into being, and demarcate space and tering and schematically synthesizing what is given to the
time through their operations. Insofar as it upholds this post- senses with what the artist already knows. Schematism is a
Bergsonian distancing from any foregrounding of space as the condition of artistic representation.

300 KAMINI VELLODI

 The resonance here with Peirce’s conception of the diagram is are no longer believed to correspond to things in themselves.
striking. Peirce defines the diagram as a schematic form of rela- Rather, objects in their appearances are understood to confirm
tions that can be ‘experimentally’ adjusted in an ongoing process to our a priori structures of representation. Kant calls this
of interpretation. When one makes experiments upon diagrams, process of confirmation—the synthesis of the intuitions of the
‘one must keep a bright lookout for unintended and unexpected sensible with the a priori concepts of the understanding
changes’.13 Thus, the logical (also rationalist) method of thinking —‘schematism’.22
begins with a hypothesis, proceeds through the construction of a Through the history of philosophical representation up to
diagram, and further moves through the process of testing the the modern moment that is given definition by Kant, repre-
diagrammatic presentation of fact with an observed reality. As sentation proceeds as a mediation of sensory data through the
such, diagrammatism is an ongoing and experimental process of synthesizing structures of the mind. When Gombrich describes
construction (reality is logically constructed; it did not already the Renaissance artist’s ‘preoccupation with structure’—evi-
exist), a process by which we attain laws with increasing certainty, denced by his appeal to pattern and model books, and his
moving from particulars to generals through an ongoing testing discovery of perspective—as reflecting our (human) ‘need for
against real (empirical) cases.14 a schema with which to grasp the infinite variety of the world
For Gombrich, too, the schema is the means by which an of change’, he invokes this Kantian outlook.23 What Gombrich
artist experimentally constructs, and continually ‘corrects’, his/ calls a schema—the egg shape, for instance—arguably func-
her interpretation. It is the process by which the artist matches tions as the (empirical—for it is not a priori; Gombrich does not
the schema he makes to the form it is to reproduce.15 Artistic argue that artists are born with these innate concepts before
production must, he argues, have a ‘starting point’, a ‘standard experience) equivalent of a Kantian a priori concept. The sche-
of comparison’, and this ‘hypothesis’ is then submitted to a matic egg-shape grounds the various instances of observed
‘process of making and matching and remaking’.16 The shape heads given in sense experience, allowing the artist to register
of an egg, for instance, may be taken as the schema for a head, his/her experience in terms of the known. All heads, no matter
the starting point for wrestling with the particular givens of the how different to one another, can be traced back to the generic
sensible in order to construct a synthesized interpretation (or form of an egg.
‘style’). The artist can then apply this style to other drawn Gombrich’s neo-Kantian view of artistic representation as
heads.17 In this way, artistic construction departs from any synthetic may be read as a response both to late nineteenth- to
notion of an unmediated relation to Nature, problematizing mid-twentieth-century developments in the psychologies of
the notion of ‘direct observation’. In fact, construction assumes perception and to the actual contemporary developments of
an autonomy that is in turn compared to an observed Nature. As artistic practice of that period. The view that art is an imitative
Peirce writes, in a similar vein, operations upon diagrams ‘take or illusory representation of nature had been sharply called
the place of the experiments upon real things’; ‘they are ques- into question by the artistic revolutions of the late nineteenth to
tions put to Nature of the relations concerned’.18 The artistic the early twentieth century. And it is from the perspective of
diagram confronts Nature, and not the other way round. this challenge that Gombrich rethinks the history of art. Whilst
Gombrich’s analysis reveals the intertwining of the history of it is questionable whether art has ever been representational in a
pictorial representation with the history of the idea of philoso- directly imitative way, in the wake of modernist and avant-
phical representation as mediating schematism. The notion that garde innovations it becomes almost impossible to uphold this
the work of art does not stand in a direct imitative relation to the notion. The most innovative aspects of the works of Paul
real object, but is rather the mediated product of a mental act Cézanne, Pablo Picasso, Henri Matisse, Paul Klee, and
corresponds, as David Summers has pointed out, to the long- Wassily Kandinsky, to name but a few, eradicate illusionism
established doctrine that ‘the immediate object of knowledge is through the exposure of artistic construction. In the after-light
an idea in the mind distinct from the external object’.19 It is this of their breakthroughs, it was no longer tenable to hold, at least
quasi-idealist notion of representation that Peirce also endorses not without a great deal of qualification, that painting imitated
when he writes that the diagram ‘takes the place of the experi- or wished to resemble nature.
ments upon real things’.20 Occasionally, Peirce goes even And yet, it was nevertheless possible to retain the view that
further, claiming that it does not matter that the Object of this art—and this is Gombrich’s claim—remains representational.
thought may be a ‘pure fiction’, since diagrammatics is chiefly First, one could claim that artists, like any of us, think in a
concerned with forms that are logically possible rather than representational way. Second, one could argue that artists are
existentially real—the forms ‘displayed before the mind’s gaze’ indebted to an inherited history of art. Whilst Gombrich
and ‘which are the chief objects of rational insight’.21 acknowledges that artists of the late eighteenth century
In their conceptions of diagrams, both Gombrich and Peirce onwards ‘struggled against the schema’ and the bondage to
reveal their debt to Immanuel Kant, who had famously argued universals, wishing instead to confront ‘the unique visual
that, for objects of sense to coalesce in a representation, a experience which can never have been prefigured and which
synthetic act of mind is necessary. Representations of things can never recur’—he argues that schematic thinking persists,

WORD & IMAGE 301

first, as it is a part of man’s habitual way of experiencing the structure of the text and the act of reading. Mitchell argues that
world and, second, since no artist creates ‘in a vacuum’ outside the only way to grasp or analyse these literary forms is through
the history of style—and the history of style is fundamentally a diagrams (which he also calls ‘images’)—schematic representa-
history of schematism. Pictures are representational since think- tions which allow us to apprehend the literary form through its
ing is, conventionally, representational and ‘no artist is ever outline and relations of parts. Building on his earlier claims on
free from convention’.24 the importance of the interaction of verbal and pictorial modes
As such, schematism is not only the means of the artist’s of representation in cultural forms, the view that texts must
representational relation to the real or Nature. It also acts as a always be treated as images, and that good literary analysis is
representational relation to art history as the history of styles. sensitive to the images which bring literary form to light,
For Gombrich, a new work, however innovative, is forever Mitchell presents diagrammatology as a way of attending to
grounded by the family of forms [the artist] has already seen, the visual nature of texts. Retaining the Peircian idea of the
by the grip of conventions and the power of traditions. As he diagram, Mitchell thus defines diagrammatology as ‘a systema-
laconically remarks (citing André Malraux): ‘art is born from tic study of the way that relationships among [the formal]
art’.25 Even if he wishes to confront ‘the unique experience that elements [of a work of literature] are represented’. In this
can never recur’, ‘the artist cannot start from scratch’; [he/she] way, diagrams connect the work of literature (and the writer)
can only ever ‘criticise his forerunners’, by adjusting the sche- with the work of the theorist/analyst, binding together textual
mas inherited from them.26 Thus, Giotto, setting out to paint construction and the interpretation/analysis of that
the human face, refers to the schema supplied to him by construction.28 Mitchell’s example is William Wordsworth’s
previous styles—in the work of his teacher Cimabue—and Prelude, a work that, Mitchell argues, invests many forms to
tests this inherited pictorial schema against his own observation convey its meaning. The conflicting forces of memory and
of a face, adjusting it as necessary. The results that Giotto action are conveyed through the wavy shape of a river. The
presents to us reveal this process of experimental modification serpentine image evokes the meandering passage through dif-
that nevertheless binds the new elements of his practice to the ferent periods of time in the author’s life, and the ‘psychologi-
continuous evolution of style. cal dialectics’ of beauty and fear, loss and recompense,
Gombrich’s view of the work of art as representational through progression and regression, activity and passivity. And, taken
its synthesis of sensory particulars with general frameworks (schema) as a ‘whole’, the Prelude evokes the form of the architectural
is consistent with his theory of art history as a development of ruin, of ‘an incomplete architectural structure in a natural
pictorial naturalism, where works of art continually augmenting landscape’. ‘The image of ruins […] is a picture of the
previous artistic ‘matchings’ of style to new given realities. It is achieved design of the Prelude.’29 Mitchell also underscores
perhaps surprising then, given the omniscience of the poststructur- Wordsworth’s own reference to diagrams (in Book 7, Part II)
alist critique of representation, to encounter echoes of this neo- as an antidote ‘for a mind beset/With images’. And even when
Kantian position in more recent appeals to the diagrammatic Wordsworth is seemingly at his most abstract and metaphysi-
within art history and aesthetics. I will briefly survey four. cal, gesturing to the pure and ideal beyond space, time and the
senses, Mitchell argues that form in his poem is always inter-
The persistence of representation: recent art twined with concrete, sensible reality. Diagrams exist at the
theoretical and art historical appeals to the point of this intertwining, and diagrammatology is the method
diagrammatic for attending to diagrams and rendering them explicit.
Recent work by WJT Mitchell, Frederik Stjernfelt, Benjamin In his compendious Diagrammatology (2007), Danish philosopher
Buchloh, and Margaret Iversen indicate the pertinence of the Frederik Stjernfelt acknowledges Mitchell’s call for a diagramma-
diagrammatic to contemporary analyses of artistic practice. On tology, but in contrast to Mitchell’s focus on the register of the
one level, these can be grasped as art theoretical responses to the verbal, he situates diagrammatics squarely within picture theory,
spatial turn. On another, they may be taken to reveal the persis- which is perhaps where we most expect to find it. He broadly
tence of Gombrich’s neo-Kantian idea of works of art as diagram- concurs with Mitchell’s understanding of diagrammatics as the
matic representations. Given the quotidian understanding of a study of the way that relationships among (formal) elements of an
diagram as a map with visual form, it is not surprising that the object are represented and interpreted by graphic constructions.30
majority of these analyses foreground visual art. Mitchell, a pio- Stjernfelt also retains the Peircian conception of the diagram as an
neering theorist of image/text relations, is unusual in setting out icon that represents, through similarity, the ‘internal structure’ of
the value of the diagram for literary theory. objects in terms of its ‘interrelated parts’, in so doing facilitating
In an important, but strangely neglected, article from 1981, ‘reasoning possibilities’ through a process of experimentation.31
Mitchell proposes a model of ‘diagrammatology’ for the ana- Thus, Stjernfelt argues that ‘the logical aspect of the picture is
lysis of literary form.27 Works of literature are, he argues, inherent in its very iconic construction, in its very diagrammatical
composed of forms that mediate between the abstract and the structure’.32 Building on a text by the Danish art historian Eric
concrete, the intellectual and the sensible, and the subjective Fischer, Stjernfelt conducts a diagrammatic analysis of an altar-
and the objective. These forms are dynamic, shifting with the piece of the Last Supper (1839–40) in the Frederiksberg in

302 KAMINI VELLODI

Copenhagen by the Danish painter C. W. Eckersburg—an rather than with general methodologies. One example is the
obscure artist who was apparently preoccupied with perspective, twentieth-century theorist and art historian Benjamin Buchloh.
penning two treatises on the subject. Investing Marcel Duchamp’s Network of Stoppages (1914) as a
Squares predominate in this painting. In a stark, box-like paradigmatic example, and describing it as the artist’s ‘first
room, the disciples are soberly seated around a cuboid table. diagrammatic painting’, Buchloh uses the category of the dia-
The figure of Christ, strongly lit and with red robes glowing, grammatic to designate a typology of drawing running through
precisely marks the compositional centre. Above his head three twentieth-century abstraction, including the work of artists such
square windows open onto a dark sky. The light does not come as Francis Picabia, Frank Stella, Andy Warhol, and Eva Hesse.38
from this natural opening, but from an unseen overhead source According to this typology, the work of art relies on (it represents,
that throws Christ, the white table surface and an unoccupied for Buchloh) an ‘externally established matrix’—a ‘readymade’
square wooden stool into sharp illumination. Fisher claims that or ‘pre-existing formal or linguistic conventions’, ‘pre-given sys-
the Last Supper enlists mathematical categories—including the tems of spatio-temporal quantification’ or ‘statistical collection of
geometry of the golden section and two simultaneous perspec- data’.39 For instance, Duchamp’s Network of Stoppages is diagram-
tives—in an iconographic interpretation of the well-known matic in its reliance on the artist’s pre-existing works, which it
passage from The Book of John.33 The work gives expression superimposes in layers. The mechanical drawing consists of
to the painter’s geometrical way of thinking. three sets of Duchamp’s earlier The Standard Stoppages (1913).
Eckersberg paints an empty stool, the apparent seat of a con- Buchloh further points out how Duchamp presents a schema
spicuously absent Judas. Its lines of construction point to a second borrowed from the technical drawings of Eadweard Muybridge
horizon lower than the horizon for the figural group of Christ and and Etienne Jules Marey, in so doing announcing ‘a graphic
the disciples. In this way, geometry acts as the ‘altarpiece’s icono- readymade’. Expelling subjectivity, and reducing drawing to a
graphical messenger’; the geometrical construction of a second ‘mimesis of mechanical, technical and commercial design’, a
horizon conveys the message that the son of perdition has moved schema of given information, Duchamp’s diagrammatism
away from the harmony of mankind as one, and is lost.34 ‘represents’ the contemporary age of ‘techno-scientific rational-
Eckersberg’s work is diagrammatic in representing, through its ity’ and its features of quantification and mechanization. For
‘internal structure’, the interrelated parts of its object. And it is the Buchloh, the diagrammatic is not just about the schematization
analysis of these spatial relations of the pictorial objects, and their of form—arguably all abstraction achieves this. Rather, it con-
similarity to the referent objects (in this case, the Biblical text) that cerns the representation of given data, the incorporation of the
Stjernfelt names ‘diagrammatology’.35 Thus, as it is for Mitchell as readymade into the domain of abstraction. In turn supplying an
well, diagrammatology pertains both to the diagrammatic logic of alternative to hegemonic modernist discourses of abstraction,
the work (its schematic representation of the painting’s forms and including that of the singularly expressive gesture. The banality
their relations, figured in this case through linear perspective) and of the diagram and its representational activity of a given field of
to the process of extracting this logic. Unlike Mitchell, however, data are starkly opposed to the modernist paradigm of creativity,
Stjernfelt understands diagrammatology not just in terms of the the inimitable and the new.
schematic form of the work, but in terms of its iconographic Drawing on Buchloh’s categories, Iversen defines the diagram as
character—the correspondence between form and content a ‘hybrid representation’ that combines abstraction from what is
being guaranteed by the work’s ‘scientific rationality’. In contrast, ‘immediately given in perception’ with the indexical. This conflates
for Mitchell, whilst Wordsworth’s Prelude may suggest correlations semiotic categories conventionally kept distinct.40 In an earlier
between form and content, this correlation is by no means an article, Iversen applies the diagram in a more coherent way, char-
exact science, and cannot be deduced methodically. In its icono- acterizing Masaccio’s Trinity in Santa Maria Novella, Florence, as ‘a
graphical precision, Eckersberg’s altarpiece invites the diagram- hierarchical diagrammatic representation of the figures repre-
matic analysis that Stjernfelt outlines. But Sternfelt believes that sented’. Masaccio frames Christ on the Cross within a vaulted
all pictures are diagrams,36 even those made within the twentieth- architectural niche, flanked by two ionic columns. The figure of
century non-objective tradition, at apparently a far remove from God stands on a ledge behind Christ, His arms touching the cross.
iconographical rules.37 The works of Kazimir Malevich, Wassily Between the heads of God and Christ is the Holy Ghost, symbolized
Kandinsky, Paul Klee, Piet Mondrian, Barnett Newman, Mark by the white dove with outstretched wings and head facing down
Rothko, and Ad Reinhart all represent something, through a towards Christ. For Iversen, this superimposition of Father, Son,
similarity that is diagrammatically produced via relations— and Holy Ghost is ‘a diagrammatic expression of the Three-in-One.
whether this be ideas or emotions, realms of ideal geometry or Hierarchical arrangement pervades representation’.41 The pictorial
even the that which-cannot-be-depicted. In this way, Stjernfelt diagram is the visual and representational schema of painted figures
produces a theory of diagrammatics with application across the and the relations between them.
history of genres and styles. Both Gombrich and Mitchell had Notwithstanding their points of distinction and the differing
done the same. problematics to which they are responding, these four examples
But some art historians have associated diagrammatics with reveal the tenacity of the commonsensical understanding of the
particular historical periods and particular artistic problematics, diagram as a representational schema. Like Gombrich, Mitchell,

WORD & IMAGE 303

Stjernfelt, Buchloh and Iversen have all moved away from a certain linguistics and semiotic theories (in A Thousand Plateaus of
understanding of the work of art as an imitative representation of 1980) and the context of an analysis of power (in his Foucault
an externally existing nature. Like Gombrich, too, they remain of 2016).45 But regardless of the specific framework of analysis,
concerned with the way works of art function as representations the basic claim remains the same: the diagram functions to
(and they all use the term ‘representation’ in their analyses) through destroy an existing state of affairs (whether those be linguistic
their own schematizing processes. Accompanying this representa- significations, over-coded ‘regimes of signs’,46 pictorial forms,
tionalism is a formalist conception of the diagrammatic work. The or images of thought) and construct a new reality (whether that
diagram presents relations between forms, and these relations con- be new sense, new figures of painting, new languages, new
tribute to the work’s internal consistency as a coherent object with relations of power, or new thought). As a generative and
‘style’ (Gombrich), ‘internal structure’ (Stjernfelt), ‘achieved design’ transformative tool, the diagram plays a crucial role in what
or ‘spatial form’ (Mitchell), the conventions of a formal readymade Deleuze calls a ‘pragmatic’ semiotics - a theory of signs orien-
(Buchloh) or an internal ‘hierarchy of figures’ (Iversen). It is this tated to life and which can function as a ‘science of reality’. As
formalistic rationale that in turn permits the work’s situation in a such, diagrams do not privilege any specific type of sign regime
historical lineage—the history of style as schematism (Gombrich), —be it visual, linguistic, verbal, or pictorial. Rather, they are
of spatial form (Mitchell) abstraction (Buchloh), or the conventions agents for the unblocking of signifying regimes and the con-
of perspective (Stjernfelt, Iversen). struction of new, polysemiotic reality immanent to the real.47
What seems to be overlooked by such conceptions of the In this way, Deleuze brings a particular slant to the dia-
diagrammatic are the disruptions and deviations of art’s material gram’s etymological roots. He takes us beyond the graphic and
work from its adherence to formal schema. Whilst Gombrich visual associations of the diagram to focus on the productivity
notes the ‘freedom of invention’ within Villard de Honnecourt’s of the interval. It is what is across writing, and across regimes of
diagrams, and the way Constable wrestles ‘with the unique visual signs, something nebulous and as yet unformed, that constitutes
experience’—that element which cannot be schematically prefi- the matter of the diagram. A dynamic outline, a process of
gured—he appears to see these moments as temporary distur- mapping rather than a mapped product, a diagram does not
bances to a rule of artistic production and a reality of human fully resemble or represent anything, but is rather a transitory
perception. Similarly, whilst Stjernfelt notes that Eckerbsrg aug- sign, germ of order, possibility of new facts whose lines traverse
ments the tradition of one-point perspective with his own ‘geome- different, intertwined registers of experience—the social, the
trical way of thinking’, for him the painter nevertheless remains linguistic, the semiotic, the aesthetic, and the psychic. In this
embedded within a continuous history of schematic thinking that way Deleuze also distinguishes the diagrammatic from the
connects him to Malevich and Kandinsky. Mitchell, too, claims to Kantian schema. Whereas the schema synthesies two hetero-
be interested ‘not only in the synchronic study of individual works, geneous forms—the form of the concept with the form of the
but also the history of forms, particularly the recurrent, conven- sensible as appearance48—the diagram is a modulator that
tional forms we call genres’.42 For these writers, the diagrammatic permits original interactions of relations of forces across inten-
is a trope that binds artistic production—in all its innovations—to sive continuums of matter. Whereas the schema synthesizes
the established givens of art history. Buchloh shares this outlook: a priori concepts that are ‘too big’ for the intuitions supplied
for him, Hesse, Warhol, Picabia, and Duchamp are linked to each by sensibility, the diagrammatic work of art replaces transcen-
other by an analogous appeal to ‘pre-established schemas’, and as dental possibility with a constructive operation ‘directly’ in the
such partake in a shared style. Here the diagrammatic is a trope real.
that permits the art historian to draw a new lineage, a new set of Whilst Gombrich, Mitchell, Stjernfelt, Buchloh, and Iversen
analogies that levels out difference and deviation. all situate the diagram within specific regime of signs—the
verbal for Mitchell, the pictorial (Gombrich; Stjernfelt), and
Deleuze: diagrams of sensation the visual (Buchloh; Iversen)—Deleuze conceives of the dia-
Deleuze’s conception of the diagram departs from the positions gram as that agent that disrupts the apparent primacy or
till now explored, by affirming difference and deviation as the hegemony of any sign-regime. Thus, the diagram is not just
material for new reality. In contrast to the conception of the the capture of visual form beneath the literary text (Mitchell),
diagram as a representational schema with form, Deleuze con- the ‘readymade’ within the visual (Buchloh), or the visual
ceptualizes the diagram as the agent of material and sensible schema within a pictorial composition (Gombrich, Stjernfelt,
disruptions to such a schema, as the capture of what Gombrich Iversen). Rather, it is the shaping of the ‘amorphous
had called that ‘unique’ element of experience. For Deleuze, continuum’49 of particle-signs, the heterogeneous, plural, mul-
the diagram ‘never functions to represent’ but rather ‘con- tidimensional, and dynamic field of semiotic material that
structs a real yet to come’.43 It has no form and substance of overspills the limits of signifying systems.
its own; rather, it is purely operational. It is a map of sensation, This conception of diagrammatics is posed as a full-blown
matter, and force, rather than of form. critique of representation. We have seen how for Gombrich,
Deleuze addresses the diagram in the context of painting (in and Peirce, diagrammatic synthesis involves already known
his Francis Bacon. Logic of Sensation [1981]),44 the context of forms that provide ‘standards of comparison’. Here, the

304 KAMINI VELLODI

creative component of production is inseparable from its his- nevertheless contains the seeds of ‘possibilities of fact’. The new
toricity, and what is already established. But Deleuze dissoci- figure of a faceless head that emerges reveals its wry abandon-
ates the genetic element of the diagram from its conventionally ment of any egg-like models. Any resemblance to what we know
referential (which is to say, representational) character. He or recognize a head to be emerge only through ‘non-resembling
writes that in its construction of ‘points of creativity’ the dia- means’, through a modulation of relations that creates ‘a sensible
gram ‘is always “prior” to history’, where history is a practice rather than a formal resemblance’, an ‘aesthetic analogy’.54
that reassigns signification and ‘standards of comparison’ to Bacon invests the diagram as a means of destroying the clichés
creative diagrammatic acts.50 Deleuze strongly opposes imposed upon the artistic process by the reservoirs of past experi-
Gombrich’s view of the artistic schema as a ground for practice ence, including the experience of art history. For him, the diagram
and the means by which artists are bound to a linear history of is the means of destroying that which Gombrich believed to con-
style. His Difference and Repetition (1968) opens with the proposi- stitute the ongoing source of pictorial production. Bacon’s paintings
tion to ‘think difference in itself independently of the forms of break with what artists have known and assumed about figuration,
representation which reduce it to the Same’.51 That is, differ- and what everyone is supposed to recognise as figurative painting.
ence must be thought ‘in-itself’, as the element of a dynamic Instead, the diagrammatic operation produces a head that defies
construction, and Deleuze argues that works of art participate representation, a head that ‘can only be sensed’. In the artist’s own
in this construction rather than in the preservation of analogies. words, it ‘unlocks areas of sensation’ instead of simply illustrating
When the work of art frees its construction of heads from egg- the object; it acts directly on the nervous system and circumvents
shaped schemas, when a head that bears no resemblance to an the ‘long diatribe of the brain’.55 Confronted by painting as a pure
egg presents itself to us, we are shocked from a habitual stupor of and intense exercise of the faculties that submits the sensible to the
thought—as recognition, as common sense—and we begin to intelligible is foreclosed. Instead, sensibility ‘finds itself before its
think. Instead of referring the difference between an egg schema own limit […] and raises itself to the level of a transcendent
and a real observed head to the former—which leaves us with exercise’, becoming the source of a thought engendered in its
only an external difference, exterior to terms—the work of art contact with real experience.56 The diagrammatic work of art
affirms the ‘intensive’ difference within the latter, that which is thus acts as the condition for what Deleuze calls ‘transcendental’
‘unique’ in the experience of it. Thus, we might end up with a or superior empiricism—an exercise of the faculties under excep-
chaotic and somewhat deviant head, but this is the possibility that tional conditions, the exceptional conditions when sensation
Deleuze invites us to embrace. It is also a possibility that modern imposes itself without the mediation of the form of representation.57
art has realized. The implications of Deleuze’s conception of the diagram
Indeed, Francis Bacon—who painted more than one deviant become clearer when we return it to the case-studies of
head—also describes his paintings as diagrammatic. He speaks Mitchell, Stjernfelt, Buchloh, and Iversen.
of the process of casting ‘involuntary’ marks, surveying ‘the From a Deleuzian perspective, the diagram of Wordsworth’s
thing like you would a sort of graph’ (translated into French poem would not consist—as it does for Mitchell—in its formal
as diagramme), and the seeing ‘within this graph the possibilities organization. Rather it would pertain to the way the material of
of all types of fact being planted’. language is used to capture the ‘swaying’ and ‘turning’ of ‘the
ravenous sea’, the ‘throwing’ of ‘the stormy waters’, the ‘yielding’
I just wipe it all over with a rag or use a brush or rub it with of the traveller, and his ‘lingering’ at the mountain’s summit. That
something or anything or throw turpentine and paint and
is, in place of Mitchell’s conception of the diagram as a schematic
everything else onto the thing to try to break the willed
representation through which we can apprehend literary form,
articulation of the image, so that the image will grow, as it
were, spontaneously and within its own structure, and not my Deleuze’s concept of the diagram permits us to attend to how the
structure.52 matter of language captures the forces constituting the affective
dimension of the poetic experience that dislodges its communica-
Bacon’s diagram is not a representational schema, or a tran- tive register. This is not a visual experience of writing—which
scription of a pre-existing external matrix or code. Rather, it is merely converts one regime of signs into another. ‘Storminess’,
(in Deleuze’s words now) an ‘immanent preparatory work’ that ‘ravenous’, ‘swaying’, ‘turning’, ‘yielding’, and ‘lingering’ are not
functions ‘directly in a matter’, an ‘operative set of asignifying images—they are sensations. In fact they are passages of sensa-
and non-representative lines and zones, line-strokes and color- tion, movements to which no specific form is adequate. Mitchell
patches’ that do not partake in an established order of commu- invites us to extract a work’s formal diagrammatic structure in
nication or system of interpretation—that is in any extant regime order to render it intelligible. But from a Deleuzian point of view,
of signs.53 The genesis of painting integrates the elimination of in the encounter with Wordsworth’s diagram we are not simply
presuppositions, not referral back to the known. In Bacon’s Self interpreters representing to ourselves the decoded reality of the
Portrait (1973), the diagram disorganizes the lower half of the face, work’s construction and its action upon us. Rather, we are forced
destroying, in the work’s genesis, any resemblance to the per- to experience the work in the sensory asignification that is part of
ceived or idealized face. It is a ‘veritable mess’, which the intervals that are part of its genesis.

WORD & IMAGE 305

 Using Deleuze’s conception of the diagram, we might consider Into the apparent tribute to the new Renaissance values of
Eckersberg’s Last Supper not in terms of how the formal composition rationality and the systematic geometrical ordering of space is
of the work corresponds to its content, but rather with respect to injected an element that renders the mystery of the Trinity a
how new relations in the material of the painting are conjured in pictorial mystery. It is this eruption of the enigmatic as a
the process of the work’s genesis and how these relations that challenge to representational clarity, and the disruption of
disrupt the possibility of the correspondence between formal com- prior formal organization with asignifying material slippages
position and content.58 Our attention might return to the three occurring during the work’s process that open the way for a
windows behind the head of Christ. Fisher noted the absence in Deleuzian problematization of art’s work. Clearly, Masaccio’s
Eckersberg’s preparatory drawing of these windows, writing that catastrophe does not resemble Bacon’s. We do not have here a
they are ‘only sketched in loosely as a kind of after-thought’ and chaotic eruption of chance marks on the painting surface, or a
‘were not integral to the geometric whole’.59 These windows—in preparatory work fully immanent to the painting process
themselves not particularly noteworthy and not called for by the (Masaccio still uses a plan made in advance). Rather, we
story—erupt within the process of painting, supplementing the encounter Masaccio’s diagram (in the Deleuzian sense) in the
prior plan, and augmenting the iconographical correspondence of alteration to the projection of a prior Brunelleschian plan
pictorial form to textual referent, and the painter’s geometrical way during the work’s process, an immanent transformation that
of thinking. They sensitize us to a thread of the irrational and liberates ‘new relations’ not originally present in the plan,
asignifying that runs through Eckersberg’s œuvre but is overlooked which do not correspond to the projected form or depicted
by Stjernfelt, with his rationalist methodology. In Stormy Weather content of the work, and which in turn generate the feeling of a
(1845), a disjunctive group of figures is affected by a storm that has pictorial ‘life’ arguably more intense than the lived perception
no impact whatsoever on the nearby trees, which one would expect grounding pictorial ‘naturalism’.64
also to be swaying violently. In A Night Scene of Panic (1836), figures on The very thing that constitutes diagrammaticity for Buchloh—
a moonlit bridge are gripped by an event beyond the picture frame. the externally established matrix, or code—is for Deleuze its antith-
With such moments, the artist catastrophizes the logic of an appar- esis. In Logic of Sensation, Deleuze opposes Bacon’s diagrammatic to
ently coherent pictorial rationality through the work of painting’s the ‘digital’ coding of geometric abstraction as exemplified by
material. For Deleuze, it is such eruptions that characterize the Kandinsky and Auguste Herbin, a coding that draws on the ‘exter-
work’s diagrammatic function, producing new reality beyond what nally established schemas’ of a ‘plastic alphabet’ (in Herbin’s case),
can be represented. or symbolic system (in Kandinsky’s)—and which merely reinstates
Such eruptions are to be found even in the most strictly orga- representation by the back door.65 For Buchloh, such coding con-
nized compositions. Masaccio’s Trinity is hailed for its practical stitutes art’s reflection of its contemporary lived reality, its self-
realization of Brunelleschi’s new system of linear perspective, and presentation in the mechanized forms that reflect the data and
the spatial organization of the work—which apparently lends codes of an information age. It is hard not to read a tone of
itself unproblematically to the analysis Iversen makes—seems resignation, if not pessimism, in this classification of the artwork
resolute. But even within such a coherent organization we as reflective both of its technological context and of an internal
encounter moments of unhinging. Other commentators have history of artforms (here, a twentieth-century lineage of abstrac-
also noted them. Samuel Edgerton has remarked that the final tion). In contrast, there is little doubt that Deleuze assumes an
composition is not simply the result of a transferred design, and affirmatory position according to which the work retains an ‘aes-
was partly done freehand. The figures of Jesus and God the thetic’ distinction from the socio-historical circumstances of its age,
Father appear, he argues, ‘as if from a slightly higher viewpoint and it is this sensory distinction from empirical givens that consti-
than the fictive vault itself’.60 There is also the famous thesis, first tutes its transformational experiential impact.
made by G. F. Kern in 1913 and often cited and developed within
subsequent scholarship, that whilst Masaccio’s projection of the
architectural interior is mathematically correct, and indicates the Conclusions
use of an ‘unambiguous ground plan’, the figures fail to show the Borne on the waves of intellectual trends, concepts can find
same mastery of perspective.61 The position of God the Father is themselves entangled within points of view that may run counter
irrational since His feet rest on a platform attached to the real wall to their most radical reaches. The diagram is one such concept.
of the chapel, whilst His hands support the arms of the cross on the Brought ashore in recent decades by the trend known as the
same plane as Mary and John.62 Thus, as Bruce Cole has pointed ‘spatial turn’, it has generated interest with all the urgency char-
out, the work ‘is not the model of rationality and clarity it appears acteristic of the moment when a trend ignites. But Deleuze’s
to be on first sight’, and in fact raises many questions—concerning conception of the diagram deviates from the common, and his-
not only the positioning of God but also the exact size and torically established, understanding of the diagrammatic sus-
proportions of the fictive room and the source of light. The tained by recent commentators—as a schema, a simplified
work is an example of ‘enigmatic composing’, and a survey of it representation of a state of affairs, which facilitates the thinking
simply ‘ends in confusion’.63 process through the visual generalization it permits.

306 KAMINI VELLODI

 I have attempted to sketch a few moments within the history of or The Diagrammatic Basis of Contemporary Architecture’, in Diagram Diaries,
this idea, an idea that has interwoven notions of artistic representa- ed. Peter Eisenman (London: Thames & Hudson, 1999), 7–25.
5– The seminal texts for ‘spatial turn’ include: Martin Heidegger, ‘The Age of
tion to philosophical representation and philosophical representa- the World Picture’, trans. William Lovitt, in The Question Concerning Technology
tion as a process of schematism. The work of Gombrich is, I feel, still and Other Essays (New York: Harper & Row, 1977), 115–54; Henri Lefebvre, The
a key point of reference to any such sketch. His voice remains one of Production of Space (London: Wiley-Blackwell, 1991); Michel de Certeau, The
the clearest and most compelling articulations of how art as dia- Practice of Everyday Life, trans. Steven Rendall (Berkeley: University of California
grammatic partakes in a logic of representation. Although he makes Press, 1984); Gaston Bachelard, La Poétique de l’espace (Paris: Presses universi-
taires de France, 1958); and Michel Foucault ‘Des espaces autres’, Architecture,
no explicit mention of the American philosopher, Gombrich brings Mouvement, Continuité 5 (October 1984): 46–49. Nelson Goodman, Languages of
the work of Peirce—who was amongst the first to give a specific Art (New York: Hackett, 1976), was a key text in the formulation of ‘picture
philosophical clarification to the notion of the diagrammatic as a theory’. W. J. T. Mitchell describes the reorientation of modern thought
schema by which thought represents to itself its constructive pro- around visual paradigms as ‘the pictorial turn’, a notion in part derived
cesses—to the question of art’s work and art-historical study. The from Ludwig Wittgenstein’s picture theory of meaning; W. J. T. Mitchell,
Picture Theory. Essays on VIsual and Verbal Representation, (Chicago: University of
analogy between their positions reveals a conception of art’s work as Chicago Press, 1995), 2, 9, 11–13. On the ‘general politics of spatiality’ in critical
a representational mode of thinking, an outlook that can be traced and theoretical work, which he characterizes as the ‘cartographic turn’, see
to Kant, who in his conception of schematism formalized the image Bruno Bosteels, ‘From Text to Territory. Guattari’s Cartographies of the
of thought as representation. Unconscious’, in Deleuze and Guattari. New Mappings in Politics, Philosophy and
Gombrich historicizes schematism. In consequence the problem Culture, ed. Eleanor Kaufman and Kevin Hon Heller (Minnesota: University of
Minnesota Press, 1998), 145–75. See also David Harvey, Spaces of Capital: towards
of representation is presented not simply as involving the work’s a Critical geography (Routledge, 2001) and Edward W. Soja, Postmodern
relation to an ‘external’ state of affairs (‘nature’) but also the work’s Geographies. The Reassertion of Space in Critical Social Theory (London: Verso,
relation to its own history, that is the history of artistic thought. 1989). For Sebastian Cobarrubias and John Pickles, the spatialization of the
Gombrich presents us with a conception of the history of style as a postmodern era displaces nineteenth-century historicism and linearized time.’
history of the artist’s schematic thought. This binding between the ‘Spacing movements: The turn to cartographies and mapping practices in
contemporary social movements’ in Barney Warf and Santa Arias, The Spatial
image of thought as representation and the work’s historicizing of its Turn - Interdisciplinary Perspectives (Abingdon: Routledge, 2009) 36–59. Warf and
processes—its connection to other works through common under- Arias point out how ‘because so many lines of thought converge on the topic of
lying principles of the diagrammatic—is implicitly, and sometimes spatiality, space is a vehicle for examining what it means to be interdisciplinary
explicitly, replayed in the readings of diagrammatism by Mitchell, or multidisciplinary’, The Spacial Turn, 1. John Mullarkey provides a good
Stjernfelt, Buchloh, and Iversen. Thus, it is not only the question of overview of the use of diagrams in twentieth-century continental philosophy;
John Mullarkey, Post-Continental Philosophy. An Outline (London: Bloomsbury,
‘formalism’, and a reassessment of its legacy that is at stake with the 2006), 157–93.
notion of diagrammatics, but of the historicity of formalism. 6– Cf. the essays collected in Deleuze and Space, ed. Ian Buchanan and Gregg
But why should it be the case that what art history and aesthetic Lambert (Edinburgh: Edinburgh University Press, 2005). Throughout this
theory foreground is the commonality between disparate works, essay, and unless otherwise specified, the nomination ‘Deleuze’ refers both
and the way they can be commonly identified in terms of an to Deleuze’s sole-authored work as well as his joint work with Guattari.
7– Cf. Penser par le diagramme. De Gilles Deleuze ã Gilles Châtelet (Saint-Denis:
operation of schematic representation of the artist’s process of Presses Universitaires Vincennes, 2004); Joachim Daniel Dupuis, Gilles
thinking? Deleuze invites us to confront the work of art’s assertion Deleuze, Félix Guattari et Gilles Châtelet: De l’expérience diagrammatique (Paris:
of its difference beyond its debt to inherited (art) history and L’Harmattan, 2012); Jacob Zbeduk, Deleuze and the Diagram: Aesthetic
established images of thinking. His conception of the diagram Threads in Visual Organisation (London: Bloomsbury, 2012); and Éric Alliez,
reframes artistic genesis in terms of a differential construction of Diagram 3000 (Words). (Ostfildern: Hatje Cantz, 2012). Published in
Conjunction with the Documenta 13 Exhibition, 100 Notes, 100
matter, a process of thought that shocks thought’s representa- Thoughts. Guattari’s theorisation of diagrammatics, which strongly
tional image through the violence of sensation, and which in informed Deleuze and Guattari’s joint philosophy of the diagram, and
turn displaces an allegiance to art history based on identity, the secondary scholarship on Guattari’s diagrammatics, would warrant a
opposition, resemblance or analogy—the types of allegiance pre- paper in itself. See Felix Guattari, The Machinic Unconscious. Essays in
sented to us by the continuous history of styles. As a ‘map of Schizoanalysis, trans. Taylor Adkins (Los Angeles: Semiotext(e), 2011);
Janell Watson, Guattari’s Diagrammatic Thought: Writing Between Lacan and
sensation’, Deleuze’s diagram offers a new concept for aesthetics Deleuze (London: Bloomsbury, 2011).
and art history. 8– Gilles Deleuze and Félix Guattari, What Is Philosophy?, trans. Hugh
Tomlinson and Graham Burchill (London: Verso, 1994), 110–13.
9– See also John Bender and Michael Marrinan, who understand diagrams
NOTES as ‘representations in general’; John Bender and Michael Marrinan, The
1– Judy Pearsall and Patrick Hanks, eds, New Oxford Dictionary of English Culture of Diagram (Stanford: Stanford University Press, 2010), 19.
(2001), s.v. 10– Henri Bergson, ‘The Idea of Duration’, in Key Writings, ed. John
2– James Hoopes (ed.), Peirce on Signs. Writings on Semiotic. (Chapell Hill & Mullarkey (London: Continuum, 2002), 49–81, at 59.
London: University of North Carolina Press, 1991), 252. 11– Ernst Gombrich, Art and Illusion: A Study in the Psychology of Pictorial
3–Ibid., 249. Representation (London: Phaidon, 1962), 129, 144.
4– On the shift ‘from drawing to the diagram’ within architectural practice 12– Ibid., 126, 247.
over the second half of the twentieth century, see R. E. Somol, ‘Dummy Text, 13– Hoopes, Peirce on Signs, 250.

WORD & IMAGE 307

14– C. S. Peirce, ‘New Elements’ in The Essential Peirce. Selected Philosophical in the experience of it close up: on approaching the picture, ‘I measure
Writings. Vol 2 (1893–1913), ed. Nathan Houser (Bloomington and distances on the picture plane with the eyes and spontaneously infer
Indianapolis: Indiana University Press, 1998), 303. information about distances between foreground figures […] I observe
15– Gombrich, Art and Illusion, 64. the striking effects of contrast […] I may construct a space, in which I
16– Ibid, 272. can imagine my body moving around; this very wandering route inside the
17– Ibid., 148. landscape has the characteristics of a diagram manipulation […]. It is, in
18– Hoopes, Peirce on Signs, 250. short, impossible to reflect or speculate upon a picture - in spontaneous
19– Summers outlines the history of this idea from Plato and Aristotle to perception or with the distance of the analyst - without conceiving of the
Galileo Galilei and Francis Bacon. He argues that the modern moment picture as a diagram, manipulating with its parts according to different
arrives with Bacon’s distinction between the ‘factual’ interpretation of rules and, so doing, retrieve new knowledge of the objects depicted, be they
nature and the subjective ‘anticipation of nature’ that inaugurates repre- concrete or abstract.’; Stjernfelt, Diagrammatology, 279.
sentation as an act of mind; David Summers, ‘Representation’, in Critical 36– Ibid., 120.
Terms for Art History, ed. Richard Schiff and Robert Nelson (Chicago: 37– Stjernfelt, Diagrammatology, 278–79.
University of Chicago Press, 2003), 3–19, at 9. 38– Benjamin Buchloh, ‘Hesse’s Endgame: Facing the Diagram’, in Eva
20– Hoopes, Peirce on Signs, 250. Hesse Drawing, ed. Catherine de Zegher (New York: Drawing Center,
21– Ibid, 252. Houston: Menil Collection, Los Angeles: Museum of Contemporary Art,
22– Immanuel Kant, The Critique of Pure Reason, trans. and ed. Paul Guyer and Minneapolis: Walker Art Center, 2007), 117–53, at 119. See also
and Allen Wood (Cambridge: Cambridge University Press, 1999), A138/ Benjamin Buchloh, ‘Painting as Diagram. Five Notes on Frank Stella’s
B177. Kant refers to intuition, perception, sensation, and cognition all as Early Painting. 1958–1959’, October, no. 143 (Winter 2013): 126–44, at 120,
species of representation; A320. On Peirce’s comparison of the diagram to 129. Buchloh’s use of the diagram as a formal trope of twentieth-century
Kant’s schema, see The Collected Papers of Charles Sanders Peirce, ed. Charles artistic practice could be seen to function as a critical extension to the
Hartshorne and Paul Weiss, 8 vols (Cambridge, MA: Harvard University lexicon of index, grid, and the tableau developed by art critics such as Yve-
Press, 1931), 2.385, 5.531. Alain Bois, Rosalind Krauss, and Briony Fer. Fer also considers the
23– Gombrich, Art and Illusion, 133. Gombrich quotes Kant’s definition of diagram as a trope of late twentieth-century art, focusing in particular on
the schema at the opening of his chapter 2, and also indicates his the work of Dan Flavin; Briony Fer, The Infinite Line (New Haven and
Kantianism in his Preface, in a remark on how he has assumed a ‘categor- London: Yale University Press, 2004), 65–85. Michael Golec refers to the
ising’ approach to perception; ibid., 23–24, 55. presence of the diagrammatic and schematic in post-war US art; Michael
24– Ibid.,149, 246n. Golec, Brillo Box Archive (Hanover, New Hampshire: Dartmouth College
25– Ibid., 20–21. Press, 2008).
26– Gombrich, Art and Illusion, 272. 39– Buchloh, Eva Hesse Drawing, 117.
27– Mitchell, ‘Diagrammatology’, 623. The article is Mitchell’s response to 40– Margaret Iversen finds this useful to describe contemporary works
Leon Surrette’s critique of W. J. T. Mitchell’s, ‘Spatial Form in Literature: which she thinks do not easily fit into any extant semiotic category (such
Toward a General Theory’, Critical Inquiry, 6, no. 3 (1980): 539–67. as Hesse’s drawings on graph paper, Felix Gonzales-Torres’s Bloodworks
28– Mitchell, ‘Diagrammatology’, 626. See also W. J. T. Mitchell, Picture series, and Amalia Pica’s Venn Diagram); Margaret Iversen, Index, Diagram,
Theory: Essays on Verbal and Visual Representation (Chicago: University of Graphic Trace, in Tate Papers, 18 (October 2012). www.tate.org.uk/
Chicago Press, 1994), 4, 7. ‘The concept of spatial form has unquestionably research/publications/tate-papers/18/index-diagram-graphic-trace.
been central to modern criticism not only of literature but of the fine arts 41– Margaret Iversen, ‘Saussure versus Peirce: Models for a Semiotics of
and of language and culture in general’; Mitchell, ‘Spatial Form in Visual Art’, in The New Art History, ed. Al Rees (London: Camden, 1986),
Literature’, 539. For Mitchell, the diagram mediates between the sensible, 82–95, at 90.
‘concrete’ image of a ‘real thing’ and the supersensible abstract realm of 42– Mitchell, Diagrammatology, 624.
‘pure form’. For him, these formal elements can be transposed between and 43– Gilles Deleuze and Félix Guattari, A Thousand Plateaus: Capitalism
taken from other traditions of imagery (reinforcing a claim running through and Schizophrenia [1980], trans. Brian Massumi (Minneapolis: University
his work that ‘all media are mixed media’, that there is no such thing as a of Minnesota Press, 1986), 141–42. As such, Deleuze’s concept of the
pure visuality, and that representations are always hybrid). As such, dia- diagram is irreducible to the numerous actual diagrams that he draws
grammatology—ostensibly Mitchell’s version of a modernism against med- throughout his works—from his diagrams of Gottfried Wilhelm
ium specificity—encourages sensitivity to the ‘points of conflict, influence Leibniz’s monadic house to the diagrams of machines in A Thousand
and mediation’ between texts and images, revealing the inherent spatial Plateaus.
logic of a text, and, it is implied, the inherent textual logic of images— 44– Gilles Deleuze, Francis Bacon. Logic of Sensation (London: Continuum, 1981).
although this reciprocity is not elaborated upon here. 45– Gilles Deleuze, Foucault. Trans. Sean Hand. (London: Bloomsbury,
29– Mitchell, ‘Diagrammatology’, 626, 631. 2016).
30– Stjernfelt states that Mitchell uses the word diagrammatology ‘in a 46– That is, ‘any specific formalisation of expression’; Deleuze and
sense not wholly unlike my own’; Frederik Stjernfelt, Diagrammatology. An Guattari, A Thousand Plateaus, 111.
Investigation on the Borderlines of Phenomenology, Ontology and Semiotics (Dordrecht: 47– Deleuze and Guattari see pragmatics not as ‘a complement to logic,
Springer, 2007), 425, n. 3. syntax, or semantics’, but as ’the fundamental element upon which all the
31– Ibid., lx. rest depend’. Deleuze and Guattari, A Thousand Plateaus, 148.
32– Ibid., 285. 48– Kant, Critique of Pure Reason, ed. Paul Guyer. Trans. Paul Guyer and
33– Erik Fisher, C. W. Eckersberg. His Mind and Times (Paris: Bilingual, 1993), 11. Eric Matthews (Cambridge: Cambridge University Press, 1998), A19/B33.
34– Ibid., 60, 67. This corresponds to the passage in the text: ‘Holy Father, 49– Deleuze and Guattari, A Thousand Plateaus, 112.
keep through thine own name those whom thou hast given me, that they 50– Ibid., 142.
may be one, as we are. While I was with them in the world, I kept them in 51– Deleuze, Difference and Repetition, xix.
thy name: those that thou gavest me I have kept, and none of them is lost, 52– David Sylvester, Interviews with Francis Bacon (London: Thames &
but the son of perdition that the scripture might be fulfilled’; John 17: 6–19. Hudson, 1987), 56, 160; Deleuze, Francis Bacon, 160.
35– Stjernfelt also talks about the diagrammatic analysis of pictures in 53– Deleuze, Francis Bacon, 101; Deleuze and Guattari, A Thousand Plateaus, 142.
terms of a Peircian experimentation with the formal relations of a work 54– Deleuze, Francis Bacon, 115–116.

308 KAMINI VELLODI

55– Sylvester, Interviews with Francis Bacon, 18, 56. 62– H. W. Janson, ‘Ground Plan and Elevation in Masaccio’s Trinity
56– Deleuze, Difference and Repetition, 140. Fresco’, in Essays in the History of Art Presented to Rudolf Wittkower
57– Ibid., 56 (New York: Phaidon, 1967), 83–89, at 83.
58– Deleuze and Guattari, A Thousand Plateaus, 144–46. 63– Bruce Cole, Masaccio and the Art of the Early Renaissance (Bloomington:
59– Fisher, C. W. Eckersberg, 71. Indiana University Press, 1980), 178–81.
60– Samuel Edgerton, The Mirror, the Window, and the Telescope: How 64– Deleuze and Guattari, What Is Philosophy?, 171.
Renaissance Linear Perspective Changed Our Vision of the Universe (Ithaca: Cornell 65– Deleuze, Francis Bacon, 103–04.
University Press, 2009), 86.
61– G. F. Kern, ‘Das Dreifaltigkeit von Santa Maria Novella: Eine per-
spektivisch-architekturgeschichtliche Studie’, Jahrbuch der Königlich-
Preussischen Kunstsammlungen 34 (1913): 36–58.

WORD & IMAGE 309