Krešimir Purgar
MODALITIES OF PICTORIAL APPEA-
RING: FUNDAMENTAL CONCEPTS
Introduction: theoretical framing of appearing
Although still used as signifiers in the uninterrupted chain of semiosis, im�
ages today mean increasingly little and even less seldom do they represent. 
The availability of digital coding led to the manner of their appearing – that 
is, the ontological level of pictorial cognition – being in terms of informa�
tion and communication more important than the iconological and semiotic 
level of pictorial cognition. Following this rather technical insight, it is neces�
sary to think anew the relationship between aesthetics and aiesthesis. The Ger�
man philosopher Martin Seel has included art, image and sensousness into 
a new kind of phenomenologically based understanding of art objects that 
he calls the aesthetics of appearing [Ästhetik des Erscheinens].1 Seel’s concept 
is designed, above all, for the sake of an aesthetic analysis of systematically 
new phenomena of beauty that in the age of the technosphere are realised no 
more as signifiers of the classic European metaphysical tradition but as visual 
phenomena that are realised aesthetically and artistically through their own 
mechanisms of sensoriness. However, what seems the most important contri�
bution of the aesthetics of appearing is precisely the omission of criteria based 
on historical and theoretical canons of beauty, as well as the relativisation of 
1 Martin Seel, Aesthetics of Appearing [Åsthetik des Erscheinens, München und Wien: 
Carl Hansen Verlag, 2000]; translated by John Farrell. Stanford: Stanford University 
Press, 2005.
195
art historical topoi. This does not mean that classical artworks have ceased to 
be peaks of the humanist tradition; it does mean that their appearing, like the 
appearing of any other object, has to be comprehended in the light of the new 
paradigms of becoming, emergence and event. Whether the object in its appear�
ing will be constituted as aesthetic object or common�or�garden thing depends 
on the observer’s capacities of intuition and imagination. Although Seel does 
not limit his research to pictorial media, the aesthetics of appearing is based on 
the modalities of pictorial appearing, but not viceversa: the modalities of picto�
rial appearing are not necessarily involved in any kind of aesthetic experience. 
Having in mind that the aesthetical does not have to be pictorial, in this article 
we will give the pictorial appearing its theoretical independence. Starting from 
the difference that exists between aesthetic object and any other kind of object, 
we will try to open the discussion about what defines the aspects of “pure” vis�
ibility on the one hand and visibility of the image on the other.
The difference between visuality and pictoriality is comparable to the differ�
ence between simple sensuousness and aesthetic perception. Seel says of this:
“In principle, anything that can be perceived sensuously can also be per�
ceived aesthetically. Among possible aesthetic objects, there are not only per�
ceivable things and their constellations, but also events and their sequences 
– in short, all states or occurrences of which we can say that we saw, heard, felt 
or otherwise sensed them. Nevertheless, the concept of aesthetic object does 
not coincide with the general concept of an object of perception, because what 
is sensuously perceivable and can therefore be the occasion of aesthetic per�
ception is not for that reason already an aesthetic object. All aesthetic objects 
are objects of intuition, but not all objects of intuition are aesthetic objects.”2
Since this kind of stance clearly indicates Seel’s polemical attitude towards 
the inheritance of metaphysical aesthetics, at the beginning of his analysis he 
also distances himself from the tradition of the analytical philosophy of art 
following on from Arthur Danto: although he is in agreement with the Ameri�
can philosopher than any object can have aesthetic qualities (that is, “it can be 
aesthetically perceived”), Seel thinks that this fact is due to the manner of the 
appearing of the concrete object in the visible world, and not to the features 
2 Martin Seel, Aesthetics of Appearing, pp. 21–22.
196
PHAINOMENA XXVI/100�101
that have been ascribed to this object in some institutional or social context. 
Seel’s theory is essentially phenomenologically determined for it is sceptical 
in equal measures about the philosophical and speculative source of aesthetic 
experience and about its conceptualisation. Accordingly, he sends aesthetic 
experience back from the domain of intellectual perception to the domain of 
sensoriness, from the activity of critical reflection to the event and phenome�
nality of appearing.3 This is in a sense a return to the original Kantian teaching 
of “disinterested pleasure”: it does not mean that the aesthetic object must not 
have any purpose other than being the object of pure aesthetic pleasure, rather 
that the human mind is capable of seeing or experiencing some object – out�
side or beyond its practical function – as an aesthetic object as well. Duchamp 
did not draw attention to the neglected beauty of the urinal nor did he reveal 
its aesthetic dimension that had been suppressed for years, just as Cézanne’s 
painting is not interesting because of the artist’s experience of nature. We ap�
preciate both, like most of the great artists of modernity, because of the change 
of paradigm of the creation of the artwork, from the individual artistic ge�
nius to public critical judgement, i.e. from work�as�object in the direction of 
observer�as�subject. 
But, how the things stand with images in general? Does (non�art) pictorial 
appearing have any other similarities with the much more exclusive Seel con�
cept of aesthetic appearing, apart from the fact that in both cases we are more 
interested in visual and sensory phenomena than sign and textual narration? 
Can pictorial experience in the time and space of the technosphere still be 
articulated as iconic difference? The main thesis of this article is that today it is 
precisely the perception of difference, or the ability to differentiate a real from 
a virtual experience of the image the place in which the drama of the real, to 
put it in Baudrillard’s terms, is played out. It seems to us that in the age when 
traditional images are increasingly less differentiated from immersive synaes�
thetic experiences (which are also partly visual phenomena, but are not images 
alone), there is a need for an equal perceptual concentration for us to recognise 
either an original art object or some simple pictorial object.4
3 Op. cit., p. 23.
4 See my article “What is not an Image (Anymore)? Iconic Difference, Immersion and 
197
KREŠIMIR PURGAR
Another German author important for our discussion, Dieter Mersch, ex�
plains that the “logic of iconic structures” is also in essence the logic of the 
perception of the difference between picture and frame, i.e., image and non�
image, the iconic thus necessarily being something like phenomenon or oc�
currence, rather than text or sign.5 He claims that the “pictorially visible” is a 
differently visible than the “nonpictorial visual”, because the picture possesses 
a distinctive material status through which a difference is produced between 
on the one hand something that is visible precisely as image and, on the other 
hand, something that is also visible, but is a mere visual phenomenon that is 
not an image. Mersch says that even the totally immersive experience of IMAX 
cinema can be considered an image for there is still a border that frames the in�
the�image from the surrounding visual.6 Although in this book, and in other 
places, I also urge that in the case of immersion there is a marginal experience 
of the image and although I agree with Mersch it is still a matter of a pictorial 
phenomenon, the question remains whether that is really because, as he says, 
in the cinema we see the frame of the screen and the seat in front of us and 
we feel the specific cinema arrangement, or whether it is actually because we 
know that cinematographic apparatus is involved, together with the traditional 
institution of the cinema, which has not changed its illusionist character since 
the beginning of the era of moving images? True, much more important for 
us than this epistemological speculation is the phenomenological insight of 
Mersch according to which it is the immersive experience of the image that 
wipes out the basis of pictorial ontology: “All technical illusionism, what can 
be called pictorial immersiveness, finds in it its dynamics and its futility”. What 
the image attempts is equivalent to a paradox: “the effacement of that which 
constitutes the viewing of an image, and thus the erasure of pictoriality as a 
medium. The logic of technological progress exists due to this telos: a medium 
that negates its own mediality”.7
Iconic Simultaneity”; Phainomena, no. 92/93, vol. XXIV, 2015. pp. 92–93, 2015.
5 Dieter Mersch “Pictorial Thinking: On the ‘Logic’ of Iconic Structures”; in: �arko 
Pai� and Krešimir Purgar (eds.) Theorizing Images; Cambridge Scholars Publishing, 
Newcaste, 2016.
6 Dieter Mersch, op. cit. pp. 163–166.
7 Op. cit., p. 166.
198
PHAINOMENA XXVI/100�101
From this it follows that the observation of the modality of pictorial ap�
pearing is the fundamental precondition for both the possibility of aesthetic 
perception and also of the perception of the image in general as phenomena 
that have an interior logic different from that of reality (or continuum of real�
ity) in which they are located as objects of perception. These modalities have 
to be precisely defined, for two fundamental reasons. Firstly, because they can 
throw a new light on the still unresolved aporias of the pictorial turn, pri�
marily that part of it that dealt with the issue of the domination of the visual 
by the textual and vice versa; and secondly, because the technosphere faces 
pictorial mediality with completely new challenges: the question arises, that 
is, how to preserve a person’s capacity for the artistic transcending of real�
ity when the experiences of pictorial representation – traditional painting and 
cinematography in the “old fashioned” 2D technique, for example – vanish in 
the digital worlds of virtuality, in which transcendence is actually no longer 
possible? If art in the pre�digital era was the only means through which it was 
possible to transgress the borders of cognition/perception and comprehend 
reality outside the framework of mere necessity, then the virtual space of some 
immersive reality makes art today equally impossible and unnecessary. As 
Martin Seel and Dieter Mersch suggest, a new strategy of art accordingly must 
be identical to the new strategy of the image: the iconic that in his text of 1978 
Gottfried Boehm could still call image-as-difference, in the epochal turn of the 
technical�scientific age has to be turned into the image-as-appearing. 
In order to set up a plausible model of universal pictorial appearing taking 
into account equally changes in the mediality of images as well as Mersch’s 
contrast of the pictorial and the visual (one of the basic features of the techno�
sphere), it is necessary to liberate images of the surplus of content inscribed, 
that is, approach them as abstract entities; I do not necessarily think here of 
pictures of abstract art, but as pictures as objects set free of culturally inherited 
aesthetic content.8 Martin Seel in The Aesthetics of Appearing sets himself a 
8 In the second part of Aesthetics of Appearing is an essay entitled “Thirteen Statements 
on the the picture” in which the German author expressly states that the problem of 
the ontology of the picture is opened up more clearly if we start off from abstract im�
ages, sinve they do not bear the burden of representation: “Every theory of the image 
has on the one hand to explain how the pictured object is linked with the pictured 
199
KREŠIMIR PURGAR
harder task for not only does he attempt to split off the factual kind of appear�
ing (constitutive for any object at all) from the concrete phenomenal appear�
ing (which is a precondition for aesthetic differentiation), but also endeavours 
within the desired aesthetic norm to set up criteria for noticing those phe�
nomena that take part only in the aesthetics of appearing. Unlike the approach 
to the aesthetic object in traditional hermeneutic disciplines, like art history 
for example, in which some object – painting or three�dimensional object – is 
ascribed artistic properties in the process of interpretation, Seel’s method as�
sumes a process of subtraction or abstraction of a multitude of the phenom�
enal features of some object and drawing attention to only those phenomena 
that are aesthetically relevant. He calls this process of aesthetic reduction of all 
those unlimited and never ultimately comprehensive phenomena that make 
up the universal facticity of some object “the simultaneous and momentary ap�
pearing of appearances”.9 In other words, something can occur or appear in a 
specific way, in some context and in a specific manner of looking, irrespective 
of all those universal and permanent features of that object according to which 
it would not actually be particularly aesthetically interesting. The appearance 
of an object is its universal factuality according to which we recognise the ob�
ject within one class of visually and haptically perceptible objects. On the other 
hand, appearing is an aesthetic operation of visual focusing and sensory com�
prehension of the object in a new status: at once liberated of the multiplicity 
of its everyday appearance and also enriched with a unique “simultaneous and 
momentary” appearing. 
In this German author, as we can see, we are faced above all with a model 
of understanding artistic objects, while our analysis refers primarily to the 
much more general concept of iconic difference. To this extent the modalities 
of pictorial appearing that I shall propose here are not entirely comparable 
depiction,, and on the other how the pictured depiction is connected with representa�
tion”. In other words, the concept of representation in any event complicates what the 
picture itself is, for it is clear that representation is above all the relation between the 
presence of what is depicted in the picture and its absent referent. For this reason, 
when we have to do with non�figurative pictures, Seel concludes that the so�called ab�
stract picture “is proved to be the most concrete and hence the paradigmatic picture” 
(see: M. Seel, Aesthetics of Appearing, pp. 161–163.)
9 Seel, op. cit., pp. 46–62.
200
PHAINOMENA XXVI/100�101
with appearing as aesthetic category described in Martin Seel, but they are 
also not general categories of appearance – the merely visual or visible. I shall 
use Erscheinen only partially in the Seel sense, primarily thinking here of the 
abstraction of the symbolic�narrative content of the image through a kind of 
phenomenological turn from the textual to the iconic substance of the artistic 
(in this case pictorial) object. I shall attempt, thus, to come closer to phenom�
enological models that invoke Seel’s “event and moment of appearing”, then 
Boehm’s “iconic difference” as well as Mersch’s “pictorially visible”, and apply 
them taking into consideration above all the consequences of Pai�’s interpreta�
tion of the technosphere, an interpretation that dramatically draws attention 
to the completely new meaning of concepts like the real, the simulated and the 
virtual.10 In addition, or primarily perhaps, my intention in the sequel is to de�
scribe four basic modalities of pictorial appearing, in order in the second part 
of the book to consider with respect to concrete examples the possibilities of a 
new theory of the image outside the essentialist�subjectivist aporias. 
1. Temporality: representational, simultaneous and reciprocal images
The most important change that digital technology has brought to the 
whole of visual culture is the totally new effect of time in the production and 
perception of images. The basic characteristic of representation, i.e. of the re�
produced image, whether painting on canvas, photography or film – is that it 
is always a visual phenomenon that has come into being only after the repre�
sented event took place. The purpose of representation and the visual arts that 
have been based on it during the several�millennia�long history of images was 
10 Among the Croatian authors, �arko Pai� has most systematically argued on the 
account of technosphere as a term that defines consequences of the technological turn 
in the contemporary universe of humanist thought. According to Pai�, technosphere 
is a new time and space of the image in which new digital construction of reality does 
not aim at creating a sort of new utopia, or new model od representing the world, 
rather technosphere is now itself this new world. For more about this see in: �arko 
Pai�, Treća zemlja. Tehnosfera i umjetnost; Litteris, Zagreb, 2015. On the contextualisa�
tion of Pai�’s concept of technosphere in connection to Bildwissenschaft and also on 
his general thoughts about this specific term see in: Krešimir Purgar, “Zero Degree of 
Representation. Art, Technique and Pictorial Appearing”; Ars Adriatica, no. 6/2016.
201
KREŠIMIR PURGAR
a t , ar  Pai� has most systematical y argued on the
precisely in enabling the division between the unfolding of the experience of 
life as temporal continuum from the experience of art as cut inside the con�
tinuum. Until the appearance of direct televisual transmissions with the help 
of video�links or somewhat later with the aid of satellites, every image could 
be only representation, that is, the image always followed the principle of tem�
poral otherness or discontinuity with respect to the relentless course of time. 
The image halted time, although it was not its main ontological property, for 
the problem of pictorial anachronism has always been a priori experienced as 
innate to what the image is – that is, halted time that has always already oc�
curred. Since not a single picture that was created before the twentieth century 
could have been simultaneous with the event it depicted, the problem of tem�
porality is seen in relation to the characteristics of the intra�pictorial depiction 
and not with respect to the source reality. Irrespective of how long a period of 
time represented in the image was concerned, every representation was onto�
logically congruent with any other.
Let us take as an example a depiction of movement in four totally distinct 
categories of representation that all present some kind or phase of movement 
but differ in terms of their media basis: Caravaggio’s Entombment of 1603; a se�
ries of photographs of a horse galloping by Edward Muybridge of 1878 and his 
proto�cinematographic device called the zoopraxiscope; then the painting of 
Marcel Duchamp, Nude Descending a Staircase of 1912, the Umberto Boccioni 
sculpture Continuity of Space of 1913 and finally the oil on canvas of Gerhard 
Richter Woman Descending a Staircase of 1965. In order to depict movement, 
each of these artists had to find a manner of how to condense the time neces�
sary for movement to be distinguished from the depiction of some static scene. 
Photography, oil on canvas and sculpture require intervention in the content 
and the stylistic complex in order for the depiction to be interpreted tempo�
rally, while for the film, the media ground alone is enough: even a very static 
film depiction will always reveal the presence of time via hardly perceptible 
shifts.11 A static represented picture does not possess any time proper to itself, 
11 Th e introductory scene from Haneke’s fi lm Caché is a good example of the treach�
erous and manipulative character of filmic time. The very static first shot in Haneke’s 
film has a double role: structural and narrative. At the structural level the static shot 
taken with a fixed camera announces the director’s manipulation of film temporality, 
202
PHAINOMENA XXVI/100�101
'
but is dependent on the temporality that it shows: a classic tableau could cap�
ture only that very moment that we do indeed see in the picture. Since in Cara�
vaggio’s Entombment we do not see what immediately preceded the moment 
represented, or what comes just after it, the condensation of time is rendered 
by the dramatic gesturality of the figures that is able merely to foreshadow the 
continuity of the act of the entombment. 
In Boccioni and Richter temporality is shown with the same ontological 
restriction of the medium (one pictorial composition and one sculptural vol�
ume) but in the stylistic innovations of these two authors it is suggested, much 
more directly, that the invisible time (time left out of the picture, not represent�
ed) before and after the moment represented nevertheless did exist. Richter 
shows this time representing the long exposure of the camera, while Boccioni 
achieved the same effect by combining an imagined series of temporal mo�
ments into a single united volume. Thus both of them, and Caravaggio and 
Duchamp as well, had to sacrifice the verisimilitude of the depiction to be able 
to present time within the static media. On the other hand, Muybridge, with 
the help of experiments with a galloping horse, showed that the new medium 
of moving images would not know such a restriction: film shows time by its 
mere nature as medium, it does not exist without the time necessary for a rapid 
interchange of a multitude of static images or frames. 
The 1964 Andy Warhol film Empire shows in a drastic manner what cin�
ematographic representation would be like without the specifically filmic con�
densation of film time through editing: a single continued frame that can theo�
retically last as long as a reel. Accordingly, cinematic time came close to or was 
totally identified with real time but – what is particularly important for this 
discussion – never ceased to be representation. What happens in the Warhol 
film is that, although almost identical to the real pictorial�temporal situation 
before the camera, it is necessarily a depiction of a time that has gone irretriev�
ably. In the film, it is true, moments of cutting can be seen when Warhol and 
his cinematographer Jonas Mekas change reels. The result of this manipulation 
a fundamental constructive determinant of the film. At the level of content, the static 
frame that extends into time adumbrates the thematic linchpin, i.e., the psychological 
effect of observing and being observed (Michael Haneke, Caché, 2005, starring Daniel 
Auteuil and Juliette Binoche). 
203
KREŠIMIR PURGAR
of real and film time is that this temporal continuity and total identification 
of the two realities is disturbed only because of the technological constraints 
of the film medium.12 Or, in other words, the technological constraint of the 
medium drew attention to its fundamentally manipulative (and accordingly, 
potentially artistic) character.
The most important question is still to come: what happens with pictorial 
representation when the technological restrictions of the classical analogue 
film or painting medium give way to the iconic simultaneity of the digital age? 
Hidden in the answer to this question is the reason why addressing the is�
sue of temporality – and not representation – is the primary task of image 
theory today. My thesis runs as follows: before the appearance of direct televi�
sion transmissions and before the satellite coverage of the whole planet we 
could reasonably consider representation as the ontological specificity of every 
pictorial depiction. In spite of its lasting validity as philosophical and phenom�
enological concept, Gottfried Boehm did not think he needed to append any 
kind of technical or scientific legitimation to his concept of iconic difference, 
legitimation that today might crucially determine the difference between im�
age and non�image, as Boehm proposes to us in his key text Die Wiederkehr 
der Bilder of 1994. The turn to the image – that is, the iconic turn – he then 
recognised in the turn to the pictorial particularly in Wittgenstein’s philosophy 
of language and Merleau�Ponty’s phenomenology of the spirit, and not in a 
turnabout from traditional representation toward other, new technologies of 
visualisation. We can find the same problem in Mitchell’s pictorial turn of the 
same year, in which the turn to the image is interpreted from the position of 
the ideological criticism of existing visual regimes, and not from the spirit of a 
12 I do not think my argument is even slightly vulnerable to Warhol’s Empire being 
shown slowed down vis�à�vis the speed of the camera used during shooting. Cinema�
tographer Jonas Mekas shot the iconic New York skyscraper with a classic film speed 
of 24 frames per second, and yet Warhol decided to show it at 16. The interventions 
included a minimum amount of editing (although editing out of sheer necessity only), 
since there is no film reel big enough to shoot visual material for an uninterrupted 
period of six and a half hours, which is how long the material shot lasts. My argument 
is based above all on Warhol’s idea of showing real physical time as totally inappropri�
ate to the manipulative character of film as art from. This idea, that is, could have been 
conveyed in an uninterrupted film of two hours, which is that a single reel permits. 
204
PHAINOMENA XXVI/100�101
t i   arg t is even slightly vulnera l  to arhol'
technique that has irrevocably and radically changed the methods of pictorial 
cognition; this also explains to us why Mitchell interprets Crary’s book Tech-
niques of the Observer as, primarily, a collection of the technological symptoms 
of modern visual culture.13 Although in his celebrated text Mitchell does not 
identify technology as the main driver of the turn to the image and does not 
think that it can radically separate the observer from his “human nature”, he 
does nevertheless give Crary that contemporary techniques of visualisation, 
like CAD, synthetic holography, flight simulators, computer animation, con�
trol of movement or multispectral sensors can contribute to the moral and 
political fear of the “loss of the human”.14
In this place we should not go too far and talk about the loss of the human, 
and yet I do think it is justified to speak about the new ontology of the image. 
If we look at visual phenomena in the framework of the technosphere, from 
the position of the technological possibilities of visualisation, then we shall 
observe that classical representation is today just one of three equal forms of 
representation: the other two are simultaneous and reciprocal images, and all 
three categories constitute the new temporal dimension of the image. Why is 
it needful to discuss a temporal dimension, and in what way does time af�
fect the nature of pictorial experience? It is essential to understand that here 
it is not a matter of the kind of time that is necessary for a specific temporal 
medium, like the digital or analogue film, or the still earlier zoopraxiscope or 
kinetoscope, to be able to produce movement, but of real time which is actu�
ally now going on within the image. When we look at any film representation, 
irrespective of the edited sequences characteristic of the narrative film, or of a 
continuous, extremely long take characteristic of experimental films (like War�
hol’s Empire), we are dealing then with a precisely determined time that has al�
ways already happened. The very idea of both filmed fiction and documentary 
faction counts on images of a time that has passed. On the other hand, there 
are increasingly more cases of the media exploitation of pictorial time that is 
actually going on, which is simultaneous to all three: 1) the person looking, 2) 
13 For more on this, see: W.J.T. Mitchell, Picture Theory, Chicago University Press, 
1994, pp. 23–25.
14 Ibid.
205
KREŠIMIR PURGAR
 this, see: W.J.T. Mitchel ,
to what is looked at, and 3) that through which it is looked at. The observer, the 
observed and the medium of observing share the same temporal continuum. 
This form of visualisation is not new and it is a bit paradoxical that at the dawn 
of the television age, in the1920s, it was the live broadcast that was cheaper and 
easier to produce than the recording of a TV show on a some kind of record�
ing media.15 Such live images are formed as visual facts at the moment they are 
broadcast and without this moment they cannot actually exist.
Such simultaneous images take on a very different character in our time 
when they are no longer used only by the entertainment industry and public 
information departments to involve us in some sporting event or convey to us 
the drama of natural disasters; rather iconic simultaneity is used to take the 
place of something much more material – physical presence. The use of drones 
in missile attacks from the air, the ability to surveil in real time every little 
piece of the planet, CCTV cameras in public spaces – all these are examples 
of simultaneous images. These images do not re-present, they present; they are 
the incarnation of time and the presentness of the event – what is there and 
what is here are together made present in the continuum of time and, unlike 
representations, without that continuum, they do not exist. A subspecies of 
simultaneous images consists of reciprocal images, which also possess all the 
features of simultaneity, with the important difference that the effect of be�
ing present is enabled for both observer and the observed. The availability of 
digital services like Skype and virtual technology like Oculus Rift will lead to 
reciprocal images in the near future providing totally immersive experience. 
In a traditional classification in which images were always representation, such 
visual experiences could no longer be considered images: in the new classifica�
tion that I propose here, they are new, specific “space/event” phenomena in the 
sphere of the visible.16 
15 For more on this, see Stefan Münker “Eyes in the Window: Intermedial Reconfi gu�
ration of TV in the Context of Digital Public Spheres”, published in: �arko Pai� and 
Krešimir Purgar (eds.) Theorizing Images, Cambridge Scholars Publishing, Newcastle, 
2016.
16 I have written more extensively on visual experience that cannot be considered an 
image in “What is not an image (anymore)? Iconic difference, immersion and iconic 
206
PHAINOMENA XXVI/100�101
: i l fi �
The criterion of temporality is a direct consequence of the new technolo�
gies of the presentation and transmission of visual information and does not 
depend on the techniques of representation or reproduction and cannot be 
correlated with individual skills of picture production in the widest sense 
(painting, drawing, cinematographic technique). And now we come to a para�
dox. Although information technology has enabled immense possibilities of 
the creation and communication with the use of simultaneous and reciprocal 
images, representational images are still the main medium of art, while the 
new technology of visual simultaneity is used principally in the production 
of non�fictional contents. Does this mean that time that has “always already 
passed” sill enables a more creative manipulation of fictional (artistic) contents 
or does it mean that the artistic imagination has not yet caught up with the 
cutting edge techniques of simultaneous and reciprocal visualisation? Or is it 
perhaps about iconic simultaneity still be experienced simply as reality, and 
not as art? It seems that the magical effect of the artistic contingency of the 
image has not faded in the slightest in the face of the omnipresent simultaneity 
of live streaming or the immanence of the image that is proper to simultaneity. 
 2. Transparency: non-transparent, transparent and immersive images
A division according to transparency tells about the semiotic and phenom�
enological agency of the pictorial surface: in other words, it tells about the kind 
of visual information we recognise on a piece of paper, on a canvas, a film or 
television screen. As we shall see a bit later, transparency is directly connected 
with referentiality, but in the case of transparency it is crucial how we see, 
while with referentiality, what we see. In principle we can say that in the tra�
ditional concept of representation, “intelligibility” is much more pronounced 
in the image than transparency. This axiom derives from the concept of the 
Renaissance picture as window onto the world and from Western culture hav�
ing taken over this principle of representation as a natural form in which to 
convey three�dimensional reality into different media of two�dimensional rep�
resentations. Since the whole of the classical tradition was founded on nature 
simultaneity in the age of screens”; op. cit. 
207
KREŠIMIR PURGAR
as model of beauty and on Aristotle’s principle of the imitation of nature as the 
objective of the poet’s art, what is transparent on the painting, i.e. what we see 
through it, is actually the natural world and human interventions in the natu�
ral world (architecture, clothing, technical aids and so on). The more elements 
from nature we are able to “recognise” in some painting, the more will we say 
this image is real and realistic; it accordingly becomes more transparent, for 
“through it” we recognise what the image depicts.
The first paradox of the principle of transparency, which does not in fact at 
all diminish the credibility of this principle, is contained in the realistic nature 
of the painting necessarily leading to the domination of pictorial content over 
pictorial phenomenon. In other words, the more the attention of the viewer is 
directed to what the painting represents, the slighter the awareness of the actu�
al phenomenon of representation will be. We might take, for example, the digi�
tal medium of moving images that can be maximally transparent or extremely 
non�transparent. We will experience a highly realistic depiction of a terrorist 
attack shot with a high resolution smart phone camera by a participant on the 
spot like a high technology snapshot, almost like a contemporary version of 
the Italian Baroque tableau, but, which is particularly important, will not rec�
ognise any artistic pretension in it. If the shot is fuzzy, dark or low�resolution, 
because it was shot from a distance with a CCTV, the observer’s attention will 
be the more focused on the formal and technical failings of the image and ac�
cordingly on its phenomenological properties. The questions that the observer 
will then ask will impinge more on the area of pictorial experience (“am I sure 
of what I see?”) than on the domain of pictorial content (“did things really hap�
pen that way?”). Non�transparency and partial transparency draw attention to 
the medium itself, because in conditions of reduced realism in the depiction 
the metaphorical window will have come between observer and world, hazy 
if a shot is unclear, and in the picture of abstract art, totally non�transparent. 
Following up Martin Seel and his “Thirteen statements on the picture” we can 
confirm that the paradox, then, consists of the pictorial experience being the 
more pronounced if the transparency is the smaller or, put still more simply, 
we shall be the more aware of the image itself and its media ground if we do not 
ask the question of what the image is presenting.
208
PHAINOMENA XXVI/100�101
The second paradox of the principle of transparency – which, as I shall en�
deavour to show, perhaps more than the previous one confirms the theoretical 
relevance of the problem of pictorial transparency – consists in any possible 
attainment of the ideal of total transparency and full immersion eliminating 
the very possibility of pictorial experience. Since, as we mentioned a little ear�
lier, transparency can be considered a person’s naturalised system of recogni�
tion of the content of a picture, it follows that transparency will be present to 
the extent to which in a natural manner we manage to recognise the realistic 
nature of pictorial content, or as Kendall Walton would say, to the extent to 
which in the picture we manage to “see the world”. This author thinks that 
between highly transparent media, like photographs, and those a little less 
transparent, like hyper�realistic paintings on canvas, there is nevertheless an 
essential difference. According to Walton, photographs possess “a bit more” 
of those features of transparency that other kinds of image lack: however, the 
differences in the degree of transparency are not enough, he thinks, for us to 
make an essential cut among the different kinds of images.17 This analytically 
oriented American philosopher invokes phenomenological insights into the 
nature of pictorial experience and concepts like difference and cut that can be 
said to be equally popular in the new image science of the European tradition, 
thinking here primarily of the theoretical work of Gottfried Boehm and Jean�
Luc Nancy.
Kendall Walton is of the opinion that the transparency of photography does 
not of itself contribute to its realism, for photography in a phenomenological 
sense is always split off from the continuum of the surface on which it lies and 
this separation is enough for each picture surface to have ontological other�
ness, irrespective of the degree of transparency: the flatness of a photograph, 
its frame, the wall on which it is hung, all these are elements that contribute to 
its flagrancy: “photographs look like what they are: photographs”.18 This claim, 
of course, does not tend to support my argument, that is, the second paradox 
of the principle of transparency; however, Walton ascribes to the transparen�
17 Kendall Walton, “Transparent Pictures. On the Nature of Photographic Realism”; 
in: K. Walton, Marvelous Images. On Values and the Arts. Oxford University Press, 
2008.
18 Op. cit., p. 83.
209
KREŠIMIR PURGAR
 t  f Photogra ic ealis ”;
cy of the photograph some features that cannot possibly be ascribed to other 
kinds of visual representations. In his earlier and better known work Mimesis 
as Make-Believe, Walton avers that the realism of mimetic representation is 
produced by a series of conventions, the primary being that certain depictions 
in some situations are considered truthful, irrespective of them being inher�
ently fictional.19 From this it derives that recognition of reality in the image 
is not rationally founded, but is the product of “make�believe”, which always 
unfolds according to set rules. According to this theory, the transparency or 
mimetic nature of realistic images from art history is not the fruit of searching 
for the perfect representation of reality, but is to do with a series of historically 
changeable rules of make�believe. Although, for example, every realistically 
done portrait is always and only a fictional substitute for the real presence of 
the painted person, in the fictional world of mimetic make�believe that person 
is really present in the image. Walton says of this: “Fictionality has turned out 
to be analogous to truth in some ways; the relation between fictionality and 
imagining parallels that between truth and belief. Imagining aims at the fic�
tional as belief aims at the true. What is true is to be believed, what is fictional 
is to be imagined”.20
Accordingly, no matter how transparent it might be, realistic painting 
belongs among the categories of imaginary world because it came into be�
ing within it, just like painterly representation. For our argument, indeed, it is 
more important what Kendall Walton says in his later text, which is that the 
transparency of the photography definitely does not belong to that category 
of visual experience to which some other kinds of mimetic representation be�
long (like painting, sculpture and so forth). Walton thinks that the degree of 
realism in some depiction does not depend on the degree of its mimesis, but 
whether the given depiction belongs to the category of imagination or the cat�
egory of truth. In photography the rules of make�believe cannot be applied 
simply because different rules of perception prevail: in the photograph we do 
not imagine we see something, rather we believe that we do.21 Although the 
19 Kendall Walton, Mimesis as Make-Believe. On the Foundations of the Representa-
tional Arts; Harvard University Press, 1993, p. 38.
20 Op. cit., p. 41.
21 Walton, “Transparent Pictures”, pp. 85–87.
210
PHAINOMENA XXVI/100�101
basic difference between photography and painting is that the former always 
shows something that does indeed exist, while painting does not necessarily 
have to present really existing objects, much more important is that the pho�
tograph has brought us a totally new way of looking at mimetic scenes; that 
kind of looking that no longer has any connection with the post�Renaissance 
endeavour for realism in painting or with the usual theories of realistic art. 
For Walton the photograph is something like a “supertransparent” medium 
through which we really do see that which the photograph shows.22
This insight is important to us because in it Walton definitely sets the prob�
lem of transparency aside from the area of representation theory and shows 
that these are two very different categories of visual experience, categories 
that depend on two incompatible theoretical premises: the phenomenology 
of looking in the case of transparency and the semiotics of the pictorial sign 
in the matter of representation. Accordingly, from Walton’s insights we might 
draw the conclusion that maximum transparency in the painting can lead to 
immersive visual experience, on condition that the technique of visualisation 
by which this is achieved enables pictorial truth in which it is possible to be�
lieve (and not be imagined, as in painting). The transparency of photography 
is just one, in truth very small, but historically verified, step in the direction 
of media strategies of transparency that today can be seen in the ever greater 
dimensions of TV screens, the huge cinema screens of IMAX, the 3D tech�
nologies, VR spectacles and so on. In the book Virtual Art. From Illusion to 
Immersion, Oliver Grau showed that a kind of “poetics of transparency” has 
always been present in the pictorial representations of Western art and has 
depended on both the scopic regimes current in individual periods and on the 
technologies of visualisation available. 
We might therefore look today at the modernist revolution of the second 
half of the 19th century and the abstract art that stemmed from it more as a 
conflict with the poetics of transparency than as a conflict with the tradition of 
realism. I derive this conclusion from the conviction that pictorial immersion, 
i.e., the merging of pictorial experience with reality is part of the historical 
process of the teleology of representation – the human need to produce im�
22 Ibid.
211
KREŠIMIR PURGAR
ages that will describe, replace or amend reality. The second paradox of the 
transparency principle, then, consists in those depictions that have managed 
the most to get away from the “image of reality” having in an ontological sense 
come closest to the “true image” – naturally, this is a matter of radically non�
transparent paintings of abstract art. 
The division of images according to transparency is aimed at the introduc�
tion of a multidisciplinary criterion capable of approaching images from the 
position of their attitudes to the reality of the technosphere, and not, as was 
the case in traditional visual disciplines, from the position of attitude to the 
concept or idea of reality. This is a crucial turn that above all takes into consid�
eration that there is no longer one reality, which is capable of being perfectly 
simulated using some exceptional painting skill or highly precise technologies 
of reproduction. It is about the classical Renaissance ideal of painterly trans�
parency not only getting into the area of technical reproducibility, as Walter 
Benjamin already observed in his celebrated essay, but above all today about 
the technological generativity of the digital code, i.e., of becoming picture as 
immaterial visual phenomenon. Abstract art, as we saw in the case of Martin 
Seel, looks after the material experience and opacity of the painting as iconic 
difference, while digital technology plunges us into areas of simulation that 
undoubtedly provide some kind of synaesthetic total experience of visuality – 
but of visuality that is no longer an image.
3. Mediality: material, imaginary and virtual images
When we speak of the division of the image from a medial ground one 
should remark right at the beginning that this is not a division according to the 
traditional kinds of media carriers, like graphic prints, photographic negatives 
and positives, newsprint, screenings and so on; rather, it is about tangibility as 
categories of material and non�material appearing. Primarily I am referring 
here to the classical Mitchell division: into the concept of picture that unites 
material, solid or three dimensional objects in which one of the surfaces serves 
as vehicles of the visual information; then to the concept of image, which re�
lates to non�material, purely optical visual sensations present in the human 
mind in the neuro�cognitive form of mental awareness or non�material picto�
212
PHAINOMENA XXVI/100�101
rial experience; finally, I am thinking that one should add to Mitchell’s division 
a third medial ground that is neither generated by indexation (leaving a trace) 
on some material support nor is a purely mental picture, but depends above 
all on electronic impulses – the kind of picture that Friedrich Kittler calls the 
“calculated image”, that is, the digital file of virtual image.23
Why would I think that the theory of pictorial appearing has to be aware 
of this radically simplified division of the medial grounds of the image? Put 
in other ways, why would I think that it is more important for contemporary 
interdisciplinary picture theory, for example, to notice the difference in medial 
ground between the mechanical and the digital picture, or the mental and the 
hologram image on the one hand, than between the graphic print and the oil 
painting on the other? The first part of the answer to this question is simple: 
above all because the theory of pictorial appearing as conceived here deals 
with the ontological issues of images and not with their artistic value or social 
function. The mediality of the image is connected with its origin as visual phe�
nomenon, and not with its value as signifier or with theories of identity, like 
semiotics, at history or gender studies. The second half of the answer is less 
connected with the academic disciplines mentioned and their criteria and so 
is more theoretically specific. In the theory of pictorial appearing, that is, as 
well as the other criteria that I list (temporality, transparency and referential�
ity), it is necessary to define more precisely the manner in which images arrive 
in the world, how they appear to us and what makes them possible. As we saw 
earlier, Martin Seel is fully aware that it is necessary to look at the problems 
of meaning, value and identity, however important in and of themselves they 
were in Western art history, not as values in relation to some aesthetic ideal, as 
is the case in traditional disciplines, rather that aesthetic value always needs re�
constituting in the space that is opened up between appearance and appearing. 
Let us recall: in Seel appearance is the universal factuality of some object, while 
appearing is an aesthetic operation that depends on the specific momentary 
relation of one and the other.
23 For the calculated image see the text in Croatian of Friedrich Kittler “Pismo i broj. 
Povijest izra�unate slike”, Europski glasnik no. 10/2005. Translated from German by 
Martina Horvat, pp. 471–484.
213
KREŠIMIR PURGAR
Pictorial appearing itself is freed of this relationship simply because we are 
here – as already stated – not interested in the aesthetic and value compo�
nents of the image, rather the way we see it, or the way it appears to us. This 
problem can be more successfully analysed with the use of one more exclusive 
theory, that which deliberately leaves out the aesthetic component of the im�
age – meaning the “logic of iconic structures” of Dieter Mersch. Very clearly 
in the footsteps of the iconic difference concept, Mersch goes much deeper 
than Boehm himself even into the problem area of pictorial experience and 
creates out of the concept of iconic difference a much more useful theoretical 
tool. Mersch, that is, presents iconic difference as a kind of “medial philosophy 
of the picture” and says that mediality of the picture cannot be derived from 
its structures of representation or from a symbolic or hermeneutic reading of 
it. Each of these dominant models obscures its specific pictorial ontology. It is 
necessary, above all, to devise and create awareness of models that distinguish 
the pictorial from the non�pictorial, and not one meaning from another mean�
ing. Most images will create a specific problem here, for by what they show 
they will draw us into them and in this way divert our attention from their 
pure media ground. Mersch thinks that the solution of this problem is not in 
the image but in the gaze: only the specific logic of the gaze can make a differ�
ence between picture as picture and picture as thing: “a gaping difference exists 
between pictoriality and the creation of visibility, which nonetheless remains 
invisible”.24 In other words, the image can be constituted in a kind of “cut” or 
“border”; this border, indeed, is not made visible by itself, but only by the gaze 
of the beholder, since iconic difference cannot become visible on or inside the 
picture. If the iconic difference, as fundamental phenomenological property of 
the image, is not visible then it means that the image is not visible per se but 
only as a multiple relationship of the possibility of the gaze, the temporality, 
transparency, mediality and referentiality of pictorial surfaces.
For the theory of pictorial appearing, and particularly for the category 
of mediality, this is particularly important and so we shall attend to it a little 
more. Mersch correctly states that the special mediality of the image cannot 
24 Dieter Mersch, “Pictorial Thinking: On the ‘Logic’ of Iconic Structures”, pp. 163–
166.
214
PHAINOMENA XXVI/100�101
be reduced to a grammatical or rhetorical mode and so semiotics, hermeneu�
tics and iconology both essentially and disciplinarily miss what should be ad�
dressed as the medial in the image. One of the most egregious examples of this 
is, according to Mersch, the figure of ekphrasis, which only emphasises the 
incommensurability of image and text, and in this way or a priori guarantees 
failure in the linguistic presentation of the visual or through a discursive analy�
sis of the image turns visual mediality into the incommensurable experience of 
text.25 However, after we have rejected ekphrasis, hermeneutics or traditional 
iconology, we have still not answered the question of what actually defines 
the mediality of the image. In my opinion it is possible to answer this ques�
tion only if we understand mediality as just one of four categories that in their 
specific interrelationship create the whole ontology of the image or picture. As 
Mersch very reasonably observes, we cannot discuss mediality in the context 
of meaning and that is why every theory of the image should in its classifica�
tion system draw attention to this. Bearing this out is the earlier mentioned 
claim of Martin Seel that abstract paintings are paradigmatic pictures because 
they do not represent anything. However, when this – actually otherwise ac�
ceptable – statement is made in this way, it turns out that one is the condition 
of the other, or, that some picture is paradigmatic precisely because it is ab�
stract. In this place I would propose a different causality: the abstract picture 
is the paradigmatic picture because in the abstract picture there is the greatest 
cut between its medial ground on the one hand and referentiality on the other. 
Or as Dieter Mensch would say, because in the case of such a kind of picture 
the most visible is “a series of fissures” between image and gaze: in other words, 
in the perception of the abstract picture a whole series of “differences, aporias 
and chiasmas which evoke varied series of ‘perforations’ occurs, and the task 
of the philosophy of the pictorial that is based on the gaze has to be commit�
ted to reconstructing the mediality of the image and the specific scopophilia it 
evokes from this inherent system of differences”.26
A special task of the theory of pictorial appearing is to differentiate the 
categories of appearing in a satisfactory manner, for as we have seen it is not 
25 Ibid, p. 167.
26 Ibid.
215
KREŠIMIR PURGAR
enough just to distinguish the medial and the semiotic/iconological, but they 
have to be put into a sustainable relationship. The “pure iconic” will never exist 
independently of other kinds of appearing and so the theory of the image has 
above all to take into account the modalities of separation and the potential 
linking of apparently incompatible ontological categories. This separation is 
suggested to us primarily by phenomenology and Husserl’s concept of inten�
tionality, but also by Mitchell’s division into visible and invisible pictures/im�
ages, to which we must certainly add a whole string of theoreticians of the 
technosphere, including Benjamin, Kittler, Bolter, Grusin, Manovich, Mas�
sumi and Pai�. The digital or the virtual image is the third media ground that 
in a theoretical sense is still largely a tabula rasa in image theory. 
4. Referentiality: non-referential/self-referential, referential, inter-refe-
rential, multi-referential and meta-referential images
The fourth basic term in pictorial appearing is referentiality. This concept is 
similar to the terminological set used by traditional visual history and visual 
theory disciplines such as art history, iconography, hermeneutics and semiot�
ics. In addition, referentiality is still people’s most direct way of being in con�
tact with picture, or at least a way that in everyday communication each one of 
us is most aware. While temporality, transparency and mediality are primarily 
concepts useful for aims of the reflexive activity of theory, i.e., while the role of 
these concepts can in its entirety be understood as a kind of splitting of theory 
off from the content of the image, the concept of referentiality in this division 
lets us encompass the image as instrumental medium with pronounced func�
tions of communication, discourse and narrative. The reasons for the theo�
retical consideration of the diverse functions of images in this context need 
explaining at once. Of all the arguments given to date it should become clear 
that inside the theory of pictorial appearing, the instrumentality of the image 
is not foregrounded, before the problem of function would necessarily take 
us back again to the essentialist and subjectivist features of images (that is, 
to ”great humanist theories” and politics of identity respectively). However, 
the instrumentality of images cannot be dismissed in an ontological analysis 
like ours because the ultimate object of this theory is to show that the instru�
216
PHAINOMENA XXVI/100�101
mentality of images, or their “lives” as objects of communication, is actually 
the consequence of three modalities of appearing that precede the referential 
functions of pictorial communication, i.e., which essentially determine how 
images will be looked at and understood.
It is possible to set up various categories of referentiality only if we also have 
in mind the classical Aristotelian tradition of imitation of nature as source of 
sensory pleasure as well as (the now equally classical) semiotic tradition. This 
latter covers, among other things, de Saussure’s concept of sign and referent, 
Barthes’ interpretation of denotation, connotation and mythologisation, then 
the semiosphere of Yuri Lotman and the concepts that have been derived from 
it like Eco’s interpretation of semiosis that later developed into Eco’s particular 
criticism of semiotic inscription of meaning into the work – overinterpretation 
[sovrainterpretazione]. However, the great semiotic tradition is not there for 
us to set up within what is primarily a phenomenological theory of pictorial 
appearing some counter�method of sign and meaning, rather it is necessary to 
us simply because signifying (or the deliberate absence of it) is the everyday 
practice of visual communications. Both these arguments are a reason why 
our ontology of pictorial appearing cannot deal with the issues of what is rep�
resented, or why, but of how this is done and what the effects of referentiality 
are for the concept of pictorial appearing itself are. The semiotic construction 
of the sign and the enchaining of meanings (which we can call semiosis or 
mythologisation, depending on the source) is a suitable model for the under�
standing of referential appearing because it shares with it the idea according 
to which one visual utterance is always related to a second utterance, that is 
the next utterance is created as a function of the predecessor or in reference 
to its predecessor. If we set off from referentiality as the degree zero of pic�
torial meaning, then each subsequent category of referentiality is established 
in some relation to this degree zero, which we can call the original meaning. 
Every other, every derived image that refers to the original image in some way 
contains its predecessor, i.e., refers to it. The only exception to this rule lies in 
non�referential or self�referential images, but this will be discussed below.
If we get away from de Saussure’s concept of the signifier, signified and sign 
and think of the sign not as an arbitrary but as a motivated point of referenti�
ality (as proposed by Barthes’ mythology and Eco’s semiosis), then the iconic 
217
KREŠIMIR PURGAR
sign can be anything that in a sufficient measure recalls or refers to some ex�
tra�pictorial reality: photographs, printed matter, pictures on the screen, in a 
word, pictures of very diverse genres and media grounds. In the modality of 
pictorial referentiality the smallest unit of meaning is that within which we 
recognise some completed relation or reference to reality, and this relation in 
practice is most often set up with individual (artistic) images, film sequences 
or a photographic snapshot. These are referential images, and from them the 
chain of semiosis (or the mythologisation of meaning) starts off; these images 
then get into more complex relations, commenting on and invoking the initial 
pictorial reference. We can best understood this if we make use of the example 
of some actual pictures and compare and contrast them to other pictures that 
might stem from or be derived from them. A photograph of Joe Rosenthal 
shot in 1945 on Iwo Jima became an iconic sign of the victory of the American 
army in the war in the Pacific. During the years it became much more than 
its instrumental function tells us, i.e., it became much more than a pictorial 
reference to the event it shows – the placing of the American flag. However, 
the original or “zero” meaning in the semiotic chain has to be reduced to mere 
referentiality, for no other meaning could have arisen if this first photographic 
snapshot, which meant at the moment it was taken only what it shows, had not 
been recorded. 
On the other hand, an advertising photograph created in 1990 for HIS Jeans 
does not just show a boy and a girl lifting the American flag, but very obviously 
gets into a dialogue with the mythic meaning of Rosenthal’s photograph taken 
a few decades earlier. The ad photograph is, undoubtedly, referential, but it is 
also, and much more so, inter-referential, for it takes much more of its mean�
ing from the mythic position of Rosenthal’s photograph than from its own 
referentiality, which is based on a mimetic depiction of four youngsters rais�
ing the American flag. Its own referentiality, then, is there only for it to take 
up the thread from some other, previous meaning. At this moment, the motif 
of raising the American flag, has already got into a multi-referential field in 
which the original (Rosenthal) sign of heroism and the ironical advertisement 
heroism open up a space in which each subsequent reference to the original 
or the ironical picture are interwoven and get “out of control”. An example of 
the multi�referential image is the photograph of Thomas Franklin taken on 
218
PHAINOMENA XXVI/100�101
September 11, 2001, on the ruins of the New York Twins, showing three fire�
men putting up the American flag in the manner of the American soldiers on 
Iwo Jima, but also in the manner of the carefree youngsters in the jeans ad. To 
which of these two photographs does Franklin’s snapshot of the firemen refer? 
The original sign of heroism or the ironical persiflage intending to call into 
question the myth of American invincibility? Although Franklin’s was prob�
ably triggered by inter�reference to Rosenthal’s iconic work, a photograph in 
the open sphere of culture necessarily refers to both predecessors, for its mean�
ing cannot avoid semiotic chaining in any direction whatsoever. And so this is 
a multi�referential picture. Multi�referentiality is a much more complex form 
of pictorial interrelations for it not only includes invocation of a vast mass 
of iconographic sources, but also because in this process it is not possible to 
exclude the action of quite often opposed ideological discourses, which also 
make up part of the referential scope of the image.
Meta-referential images are what W.J.T. Mitchell calls metapictures: pic�
tures that refer to themselves, that is, to several levels of their own ontological 
position: 1) they reveal the way in which they are made or to the mechanism of 
the production of pictorial meaning in general; 2) they depict a kind of “theory 
of images” without getting outside their own pictorial medium, i.e., raise the 
question of whether it is possible to speak about images without ekphrasis, 
without language as verbal substitute; 3) they reveal the essential pictorial na�
ture: the image, that is, to be able to function at all as medium of communi�
cation, cannot be equated with reality, but not with itself either.27 Mitchell’s 
concept of metapicture (or pictorial meta�referentiality in our case) raises in 
principle the question of whether images can discuss themselves, instead of 
the traditional disciplines of iconology and semiotics doing it for them. When 
pictures are directed toward themselves and when they reveal the models 
of production of all other images, they become subversive mechanisms that 
uncover institutions and the discursive production of power. One of the best 
known examples of such a kind of painting is Diego Velázquez’s Las Meninas, 
which, with a complex system of inter�, multi� and meta�referentiality tests out 
219
KREŠIMIR PURGAR
27 For more detail about this see Mitchell’s text “Metapictures” published in W.J.T. 
Mitchell, Picture Theory, University of Chicago Press, 1994.
the status of the iconic in general as mechanism of the political production of 
power.
It is interesting that Mitchell should say that metapictures are at the same 
time radically self�directed, i.e. they create their own inter�pictorial theory of 
the image, but they are also intertextual, that is, they create meaning by being 
enchained with other pictorial utterances that co�exist in parallel within the 
visual culture of some community or historical period. If we take into con�
sideration Mitchell’s reasoning and his concept of referentiality that I am en�
deavouring to defend here, it follows that metapictures, or meta�referential 
pictures, go back to the very beginning of pictorial referentiality, i.e. to self-ref-
erentiality, or radical orientation of the images to itself. Paradoxically, images 
that are most oriented to themselves, as we saw above in the case of Martin 
Seel, are those that refer to nothing except themselves, that is, pictures of ab�
stract art. I think that we might approach this paradox in the following way: 
non�referential, or self�referential images on the one hand and metareferential 
on the other only confirm in different ways the same basic pictorial ontology 
that Boehm calls iconic difference. Both kinds of images ultimately reveal their 
position of otherness as compared to extra�pictorial reality: the first does this 
by emphasising the differences between image and world, and the other by 
radical deconstruction of the way in which it is culturally produced. 
Every picture that appears in some referential modality – irrespective of 
whether it relates to something outside itself (referentiality) or is related only 
to itself (non�referentiality and self�referentiality) can in principle possess only 
the two earlier mentioned temporal modalities – representationality and sim�
ultaneity. In a pictorial�ontological sense it is irrelevant whether we look at a 
direct transmission (iconic simultaneity) of a picture that represents some�
thing or reminds us of something (referential, then) or whether we look at 
non�referential, i.e. abstract, images (for example, when in a live broadcast 
of the opening of an exhibition we look close up at some abstract picture). This 
latter picture would be simultaneous and non�referential. However, we cannot 
term pictorial experience the rather improbable, if possible, situation in which 
in front of two computer screens with web cameras turned on there are two non�
referential pictures “looking” at each other. This hypothetical situation confirms 
that the effect of referentiality is independent of the effect of temporality, but 
220
PHAINOMENA XXVI/100�101
only if the condition that Husserl calls “pictorial consciousness” is satisfied. In 
other words, an image, in order to appear at all, irrespective of the degree of its 
own referentiality, has to be “produced” by the gaze. The need for pictorial con�
sciousness to exist tells us that this fundamental phenomenological insight sets 
up a border behind which perception of the image is no longer possible. 
On the other hand, although we have seen that an image does not exist 
without the intentional consciousness (of the observer) that produces it, Sartre 
says that these two – perception and the image – are not one and the same. 
From this it derives that what makes the act of perception inseparable from the 
object of perception is just the product of consciousness that agrees to the cog�
nitive convention that Sartre calls “the illusion of immanence”.27 The illusion 
of immanence is necessary in a communication system in which conscious�
ness operates at a level different from that of physical objects, even when these 
objects are in a certain way incorporated into consciousness and constitute a 
continuum with it. The illusion of immanence enables the continuum not to 
be revealed, enables the images to be received as if they were what they show, 
although they are not what they show, but are simply images.28 The modalities 
of pictorial appearing open up the possibility of a kind of “transitional” theory 
of images for contemporary times that are not (only) analog any more while 
not being (only) digital either. This theory takes into account the vast space 
that still exists between objects and perception, that is, between eikon and pure 
sensousness.
28 See Jean�Paul Sartre, The Imaginary. A Phenomenological Psychology of the Imagi-
nation, Routledge, London 2004. [1940], p. 12; then John Lechte, “Some Fallacies and 
Truths Concerning the Image in Old and New Media”, Journal of Visual Culture, vol. 
10, no. 3, 2011, pp. 357–358; and in Krešimir Purgar, “What is not an image (any�
more)?”, op. cit., pp. 166–167. 
29 Jean�Paul Sartre, op. cit., p. 6.
221
KREŠIMIR PURGAR
t ,