mimesis and semiosis

an inquiry into the relation between image and sign


There is a difference between image and sign, There are many kinds of signs according to the degree of motivation, and different kinds of images according to the degree of schematisation. These differences do not entail a transition from image to sign or from sign to image. There are many ways in which image and sign work together: the many forms of entwining of semiosis and mimesis. The difference between image and sign sheds a new light on the difference between art and science or philosophy.


Mimesis - the making of images of an original - is a phenomenon that can be understood without explicit reference to semiosis. But it is important to distinguish it clearly from semiosis - the referring to a meaning through a sign. As a matter of fact, the image is often understood in terms of the sign - from way back, but especially in the twentieth century (see for instance Nelson Goodman, Arthur Danto, W.J.T. Mitchell, Gottfried Boehm, Lambert Wiesing). A clear understanding of mimesis is impossible, hence, without distinguishing it properly from semiosis. Add to this that precisely a clear insight into the nature of signs makes it understandable how something like mimesis is possible at all, as will become evident in the course of this text.

Two good reasons to examine the relation between mimesis and semiosis.


Let us begin with a brief description.

We are dealing with an image, when we perceive something where there is in fact either something else (for instance a canvas) or nothing at all (dream images). The image is determined by similarity as well as by difference. To the eye, the image in the mirror does not differ from the original. That is why we take it for the original. But the other senses can descry the difference at any time. That is why we experience the image as a mere imitation. An image - an imitation - is not a copy, hence, not a clone, no serial product, but a sensory reduced version of an original.

Totally different is the sign. A sign is a sensory configuration that functions as a substitute for something else - an object, and idea, a state of affairs, and so on - which is the referent or the meaning. The refinements contained in the terms of diverse semiotic triangles do not concern us here: as a matter of convenience, we will simply talk of 'sign' and 'meaning'.


An image is characterised by its relation to an original, the sign by its relation to a meaning.

Both relations are often assimilated: an image would refer to its original like a sign refers to its meaning. That would come down to a subsumption of the image under the sign, and the corollary transformation of the image into a particular kind of sign. Such assimilation is unjustifiable, as becomes evident on closer examination.

Before proceeding, we have to make a terminological remark. In English, the term 'to represent' means 'to stand for' as well as 'to depict.' In matters of mimesis, this is the source of much confusion. A sign represents a meaning in the sense of 'standing for', not in the sense of 'depicting'. In the following, we will use the term only in the sense of 'standing for'.

Making an image makes only sense when the original is absent - elsewhere in time or space - although we would like to contemplate it nevertheless. By making an image of it, we make it accessible here and now: the image presentifies an absent or non existent original. A sign, on the other hand, makes an absent state of affairs here and now accessible in providing a substitute perception. When someone says to me 'it rains', I go outside to fetch my things. It is obvious that these words do not presentify what is absent - I do not have the impression that I grow wet - but represent it ('stand for it') through a substitute perception - the words 'it rains', which refer to the rain that makes the things outside wet. Precisely because the substitute perception is not the thing itself does it refer to its meaning. Referring is the countermove elicited by the initial move contained in representing: representing implies referring. Otherwise than with the relation between image and original, which is always a spatiotemporal relation between something given here and now and something that exists elsewhere in space or time, the referring relation between sign and meaning is not always a spatiotemporal relation. Only exceptionally is the meaning something that is perceptible, like the rain outside: as a rule, we are dealing with imperceptible states of affairs, that are not situated in time or space - just think of concepts like 'image' and 'sign'. Only the signs for these meanings are perceptible here and now, so that we can only metaphorically speak of 'referring'. Only when it is agreed that referring can be understood in a metaphorical sense as well, can we state: the image presentifies, whereas the sign refers because it represents (in the sense of 'to stand for').

the image


an original
the sign

refers to

a meaning


The difference between presentifying and referring is closely related to another one: the perception of a sign is purely instrumental, whereas that of an image is a goal in itself. The image is something that wants to be contemplated: we do not react to it otherwise than in continuing to look at it. Things are totally different with a sign: as soon as it is recognised, we stop contemplating it. Instead, we proceed to reacting to it. When we see thunderclouds, we look for a shelter. When we see a red traffic light, we stop. When we hear somebody shouting 'Fire!', we try to escape from the building. When someone asks us the way, we show him the way. When someone contends that the image is a sign, we react by trying to test the contention and to confirm or deny it. Thus, the sign prompts us to some reaction. That is why we look at a sign only as long as is necessary to comprehend it - and thereto a glimpse amply suffices, exemplary in reading a text. That explains why the sign is so easily reduced to a ghostly scheme, whereas the image, by nature, tends to be saturated.

For a better understanding of the difference, it pays to reformulate the problem in the broader context of perception as such. It appears that there are two kinds of perceptions. To begin with, there are the perceptions of which we want that they endure: pleasurable perceptions (in the broad sense of the word: it may be pleasurable to see the guilty suffer, and so on...). Far more frequent are perceptions that we want to avoid (hunger, pain), or that are only means to avoid other unpleasurable perceptions or to realise pleasurable ones. The real objects or situations that are read as signs, can thereby be replaced with verbal signs that refer to them: instead of reacting to the smoke as a sign of fire, we can react to the call 'Fire!'. In all these cases, signs prompt us to produce a reaction, that in its turn produces another sign that prompts us to another reaction, and so on. The chain is prolonged until eventually every perception becomes obsolete either because there are no more perceptions to be avoided, or because we have reached a pleasurable perception to enjoy. The chain may be interrupted when other chains are more urgent. The reaction is then postponed. This can go so far that no reaction follows at all, as is the rule with general information or insights from scientific texts. The reaction is then suspended. Also when the reaction is postponed or suspended, we do not continue to contemplate the sign: the information is stored in our memory for possible later use. Thus, the sign is a shackle in a chain of perceptions that annihilate one other. The pleasurable perception, on the other hand, is an end station where we would like to stay forever. We can render this state of affairs in a scheme as follows (s= sign and R = reaction):


sign s1 → R → s2 → R → s3 → R or


This model shows the opposition between the chain of signs whose perception is annihilated by reactions on the one hand, and pleasure, as a perception that perpetuates itself, on the other hand. The sign as an instrumental perception is opposed to either the absence of perception or a pleasurable perception. Where does the image come in?

What is pleasurable is not always present here and now: it has gone by, has still to occur, is elsewhere, or does simply not exist. Making an image of it is an appropriate means of making it present nevertheless:


sign s1 → R → s2 → R → s3 → R or



That the sign is merely an instrumental shackle in a chain of reactions, whereas the image is an end in itself, provides a excellent criterion for discerning an image from a sign. It suffices to ascertain whether we continue looking at it, or whether we make the perception of it obsolete by performing a reaction: we only take a furtive glance at a traffic sign, but continue looking at the Mona Lisa; our eye dwells upon the gracious figures of Adam and Eve, but we take only a furtive glance at the pictogram below (an imagesign, as we shall soon see) before entering the appropriate door:

adam and eve


That we cannot stop looking at images, has everything to do with the kind of originals it shows: no mere instrumental perceptions - no signs - but rather pleasurable originals (in the broad sense of the word).

Nothing prevents signs form having a strong emotional charge - which only highlights their substitute character: whereas it is only understandable that many a Greek left a stain on the marble after his nocturnal visit of the statue of a nude, the irrationality of burning a flag catches the eye.


As opposed to our claim that there is a fundamental difference between image and sign, the idea is widespread that there is something like a continuum between both. Two factors seem to favour such a claim: from the point of view of the sign the existence of different kinds of motivation, and from the point of view of the image the extreme variety of kinds of images. Let us first examine whether an increasing degree of motivation justifies the concept of a transition from sign to image. Let us therefore first explain what is meant by the motivation of a sign.

The relation between sign and meaning can be totally arbitrary - there is no inherent relation whatsoever between the sign and its meaning. That is the case with 'unmotivated signs' like words - the signs par excellence. In other cases, there is some relation between the sign and its meaning - analogy, metonymy or similarity. That is the case with 'motivated signs'.

There is not only a difference between sign and image with regard to the 'spatial' relation between original/imitation and sign/meaning (referring versus presentifying), but above all with regard to the motivation of the relation. Unmotivated signs - the signs par excellence - are the complete opposite of the image. Whereas, put aside the sensory reduction, the image is identical with the original, there is no relation whatsoever between the unmotivated sign and its meaning. As the degree of motivation increases, the sign comes gradually to resemble the image, so that the impression grows that there is a continuum between unmotivated signs and images. To be sure, there are various degrees of motivation, including motivation by sensory reduced similarity, or even complete similarity. But whatever the kind of motivation, we are always talking about the relation between sign and meaning, never about the relation between original and imitation.
Whereas the relation between sign and meaning may vary, the relation between image and original is always determined by (sensory reduced) similarity. And whatever the nature of the motivation of a sign, it always remains a sign, and never is transformed into an image. It pays to dwell somewhat longer on this subject. Let us first give a concise survey.

In their purest form (case 1) signs are unmotivated. The unmotivated signs par excellence are those of verbal language. Combinations of phonemes are arbitrarily assigned a given meaning: there is no inherent relation whatsoever between 'ball' and 'bill' and the corresponding meaning.

As opposed to unmotivated signs, there are motivated signs, which are related to their meaning in one respect or another. For our subject, it is important to distinguish four kinds of motivation:

To begin with, there is motivation by analogy (case 2): perceptions for another sense, or imperceptible phenomena are translated in perceptions of non-existent, arbitrary configurations for a given sense (for instance: a curve). Most frequent are visualisations of non-visual givens (warmth, pitch) or of sensory imperceptible givens (infrared, ultraviolet, radiation, quantities and what have you). According to the sense addressed, we can speak of visualisations, sonifications and so on.

Next, there is motivation by metonymy (case 3): an abstract idea can be replaced with an instance of it (as when a dove refers to 'peace'); or a spatiotemporal complex of perceptible phenomena can be replaced with a part of it (as when someone's chair is used to portray him, or when corpses refer to a slaughter). Whereas with visualisation, the meaning is translated in an arbitrary sensory configuration, with metonymy, it is replaced with (the image of) an existing object. We propose to call this kind of sign a symbol.

Next, there is motivation by (sensory reduced) similarity (case 4): an image is used, not so much to presentify the absent original, but eventually to refer to it. The image is then used as a sign. Examples are the images of male/female on toilets, or the images of dishes in a restaurant. There are many terms for this kind of signs (icons, pictograms....). The term 'icon' comprises not only symbols, but also the image as such. Therefore, we prefer to speak of imagesigns (Case 4).

Finally, there is also motivation by complete similarity (case 5): a thing is used to refer to a copy, another product from a series, or a clone. That is the case with items in the shop-window that refer to other instances of the series in the shop. We propose to speak of objectsigns.

Let us give an overview:

sensory nature
of the meaning

nature of the transformation
into perception

nature of the sign

nature of the relation
between sign and meaning

kind of sign




thing itself

complete similarity


visual similarity
sensory reduced similarity


non-visual or

visualised into

(image of) object




non-visual or

visualised into

visual configuration

sensory analogy




visualised into

sensory configuration



Let us examine each of these cases:


The paradigm of the unmotivated sign is verbal language. Phonemes - sounds, organised in a system of oppositions - are combined into morphemes and words which are arbitrarily assigned to meanings: there is no inherent relation between 'bill' or 'ball' and the corresponding meanings.

The auditory signs of language come to be signified in their turn by visual signs: writing. Just as verbal language consists of a system of audible oppositions (phonemes), the alphabet consists of a system of visual configurations (lines) combined into letters, which are concatenated into words, lines, pages, books:

p q n u v


Comparable with the alphabet is the code of numerals: Roman or Arabic numerals can be combined into complex numbers, where also the place has a meaning. Next to codes of numerals, there are also codes of mathematic or algebraic signs (+, -, x, %, ...).

Al these signs are more or less like writing, but there are also signs that are more like images (with two-dimensional configurations): think of colour codes or the codes of flags consisting of oppositions between 'optemes': colours, horizontal and vertical planes, relative position and so on. Combinations of such 'optemes' are conventionally assigned to certain countries: there is no reason why a vertical disposition of blue, white and red would designate 'France'. Below, we render a part of the code:

Although these visual configurations are more similar to 'images' than to 'writing', we certainly cannot talk here of images - think of abstract paintings - in the literal sense of the word.

Needless to mention that there are also auditory codes alarm calls, military signals, ringtones, and what have you.


A visualisation is the translation of perceptions of other senses or of phenomena that are imperceptible altogether into an appropriate arbitrary configuration for a given sense.

Examples of translation from one sense into another are a temperature curve, where changes in temperature are rendered visually as changes in the 'height' of a line on a two-dimensional plane, a weather map where different colours render different temperature zones, or a score, where pitch is rendered visually in 'higher' or 'lower' positions on a staff. There are also scores with other visual analogies (with free interpretation code) like those of Cardew:

cardew score

The most simple visualisations consist of single translations: the height of a column of quicksilver, the wind direction on a weathercock. Such single translations can be composed into complex databases - diagrams:

temperature curve

Examples of translation of imperceptible phenomena in sensory perceptible configurations are photo's of space, where invisible radiation of all kinds are rendered in colour, scans of brain activity, relief maps where differences in altitude are translated into colours, columns that represent (stand for) quantities (diagrams), Venn diagrams, and what have you. Visualisations can be three-dimensional: just think of models of atoms or DNA-chains.

It is obvious that such visual configurations are not images: what is not visible or not perceptible at all, cannot simply by imitated visually - which is precisely why we resort to motivation by analogy. There is only some superficial likeness to (some aspects of) images: a temperature curve has something of a drawing and a temperature map something of an abstract painting. That does not prevent many authors from regarding maps as images (Gottfried Boehm) or visual scores as instances of the combination of visual and auditory images (music)...

Let us remark that visualisations, just like imagesigns, can be approached like images. We can read a temperature curve as a drawing of mountains, like Nelson Goodman, who cannot properly tell a temperature curve from the rendering of the Fujiyama by Hokusai. But such approach only emphasises that a visualisation is not an image.


Next, there are the two cases that are coloured in grey in the above table: the field of the images that are signs.

In the case of the symbol, an image is used as a sign for the (spatially and/or temporally more complex) whole of which it is part, or for an imperceptible idea. Examples of the former case are: the bowler hat and the cane for Charlie Chaplin, the White House for the government of the United States, cheese for the Netherlands, a young woman for the country, smoke for fire, the corpses for the massacre, and so on. Examples of the latter are hieroglyphs, the image of a dove as a symbol for peace, the image of a hyena as a symbol for greed, and what have you. We are dealing here with what in rhetoric is called metonymy (in the broad sense of the word: synecdoche included, metaphor excluded as we shall see soon).

Symbols can be part of more encompassing statements. That is the case with symbolic predication: as when we depict someone with a dove in his hand to convey that he is bringing peace, or a person with a palm to indicate that he is a martyr, or when someone is shown in front of skeletons to contend that he treats people as mere cannon fodder:

heartfield bismarck

Equally cherished is the metaphor: the ascribing of a predicate to something by replacing (part of) it with the symbol of the predicate. An example of a verbal metaphor is calling Rommel 'the desert fox'. An example of a pictorial metaphor is Heartfield's statement in which the body of a capitalist (symbolised by a hat) is replaced with that of a hyena. Equally from Heartfield is the replacement of the arms of the swastika with axes or the snake around the shoulders of a woman in 'Sünde' from Franz von Stück:

von stück sünde

heartfield nazi

Finally, single symbols can be added into an allegory. Examples from literature are Swift's 'Gulliver's Travels', Kafka's 'Das Schloss' or Bertolt Brecht's 'Mahagonny'. Examples from the plastic arts are the many allegories of Bosch and Brueghel, or Marcel Duchamp's 'La Mariée mise à nu par ses célibataires, même'.

bosch hooiwagen
duchamp glass

An example from philosophy is Plato's allegory of the cave. In all these cases, we are dealing with images that are signs. The image presentifies the dove, not to show us an absent dove, but rather to refer to peace. It is not the intention that we admire an absent dove, but rather that we insert it as a non-verbal sign in a statement about peace. That applies equally to the figures on Botticelli's Primavera, or for the hey wagon in Brueghel that is part of the statement 'The world is a haystack and everyone grabs what he can'. We are dealing, hence, with images that are not considered for themselves, but rather refer to a more complex whole or an abstract idea. That is why we are speaking of instrumental mimesis.

Let us remark, again, that also symbols can be looked at like images as such, just like we can read 's' as a snake, or a temperature curve as a mountain chain. They are no longer signs then, and what they mean is no longer of any concern.

Pictorial metaphors or symbols are easily recognised in that they tend to look 'surrealistic'. their coherence is not determined by what they depict, but by what they mean. In literature, the border between image and symbols is not so apparent as in the plastic arts, but therefore no less real.


An image can also be used to refer to its original: the image is no longer a symbol then, but an imagesign.

Images can be used to identify someone, to document an accident, to sell a product, to show what havoc has been wreaked by an attack, and so on:


twin towers

Also three-dimensional images can refer to real objects: models of cells, organs, dinosaurs, primeval men, buildings, airplanes, and what have you:

opera paris maquette
heart model
anatomic model

As we shall soon see, imagesigns are not always true to nature: many pictograms are schematised - think of emoticons.

Only seemingly are we dealing here with images: they presentify the original only in view of a subsequent reference to it, just like a sign. The image is only a means of referring to the real world. Just like with symbols, we are dealing here with mere instrumental mimesis.

Just like visualisations or symbols, also imagesigns can be looked at as if it were images. That would be the case if we would regard da Vinci's drawing of copulation below as the encounter of two imaginary beings consisting of a conglomerate of intestines and cut-through spinal columns:


Not only images can be used as sign for the original, also real mass-produced articles can be used to refer to other products of the series. Such objectsigns demonstrate clearly that there is no more a transition from sign to image than that there would be a transition from the sign to the real world.


There is an irresistible propensity to provide unmotivated signs with some motivation. The endeavour may be limited to attempts to discover some rationality in the choice of the colours of the flag (black = earth, yellow = corn, red = blood of the heroes). In other cases, the action is suited to the word, as when someone writes big as 'BIG'. Such propensity betrays the need to transform the ghostly sign into a full-blooded image - if not into the thing itself. Conversely, increasing motivation can make signs unfit to convey a precise meaning, so that they have to be combined with less motivated signs. Thus it comes that there are more than five kinds of signs according to motivation: they can be combined with one another in new units like onomatopoeias and route maps. We give a extensive overview in an attachment 'Combination of signs with different degrees of motivation' (Return with the link at the end of the attachment).

Especially in the cases where we are dealing with condensation, it seems as if there is a transition from one kind of motivation into another - as if the word would become image, like in:


The most dramatic example is the additional motivation of visualisations with imagesigns, as in road maps. But there is never a gradual transition from one kind of motivation into another (in case: from visualisation to image sign), let alone from a transformation of sign into image: there is only a different mix of clearly distinguishable parts. It is revealing that the additional motivation can never be sustained systematically: in every language, only rare objects are designated by onomatopoeia, not all the objects that produce sounds. Even when the replacement can be systematic - as when, in a written text, all the substantives are replaced with images, or when in a spoken text all the substantives are replaced with the sound they produce - the images continue to function as an imagesign.

The fact that symbols and imagesign can be read as signs - and conversely - does not suffice to assimilate the image with the sign - be it a symbol or an imagesign. It is not because you can use your laptop as a weapon, that there would be no difference between weapons and computers.


Our inquiry into the motivation of signs shows that there is a fundamental opposition between the image and the sign. To be sure, signs may be motivated by (sensory reduced) similarity (imagesign, symbol), and even by complete similarity (objectsign). On top of that, images with different kind of motivation can be combined with one another. But the degree of motivation has no effect whatsoever on the status of the sign: a sign continues to refer to an absent meaning, it never presentifies it. The reason is that the degree of motivation has everything to do with the relation between sign and meaning, and nothing with the relation between original and image, which is invariably determined by sensory reduced similarity:

relation original/ image

relation sign/ meaning

visual similarity

image (portrait)

- - - -
visual similarity
visual similarity
image sign (photo hamburger)
symbol (pictogram with knife and fork)
- - - -
- - - -

visualisation (temperature curve)
spoken word

sensory analogy

The effect of variations in kinds of images and variations in kinds of motivations is analysed in detail in 'Visualisation: map and landscape'. (Return to this place with the link below the analysis).

The degree of motivation does not determine a continuum between sign and image, as W.J.T. Mitchell (with his imagetext) and Gottfried Boehm contend.


We have seen that there are two factors that seem to support the idea of a continuum between sign and image: from the point of view of the sign the existence of different kinds of motivation, and from the point of view of the image the extreme variety of kinds of images. Many authors point to the fact that there is a continuum between images that are more 'saturated' and images that are more 'schematic', and that this continuum coincides with a continuum between image and sign. After having demonstrated that the degree of motivation does not allow to conclude to the existence of a continuum, we will demonstrate that existence of different kinds of images does not suffice either.

In our text on mimesis, we introduced three variables that allow to account for the differences in kinds of images: the image may be a more or less suggestive rendering of the original, and the original may differ from the model by idealisation and essentialisation. Below is a table were the degree of suggestion increases from the left to the right, and the degree of idealisation and essentialisation from above to below:

michelangelo david

da vinci foetus
matisse zulma
brancusi torso
matisse line drawing

These examples show that neither the degree of suggestion nor idealisation/essentialisation does transform an image into a sign: in al the above cases, we are clearly dealing with genuine images.

That is equally the case with images that function as signs. With the imagesigns below, the rendering of the blood vessels varies from more to less suggestive: the flat rendering on the imagesign on the left and in the middle are a suggestive two-dimensional rendering of a three-dimensional original. Only in the imagesign on the right is the rendering non-suggestive (shadows), so that we can better situate the vessels in the third dimension.


And that applies equally to essentialisation and idealisation. A good example of essentialisation is the reduction of the body to an isolated system: vascular system, digestive system, muscalatory system, nerve system and so on. Only the skeletal system can easily be essentialised. An isolated vascular system, on the other hand, is only conceivable as in image. In the three imagesigns above, we see three degrees of essentialisation. To the left, the entire vascular system is reduced to the two (double) circuits of blood. In the second and third imagesign, the degree of essentialisation is reduced, and the circuits are rendered in more detail, the relative importance of the vessels is evident from their width, and we learn more about their spatial distribution. In the second imagesign, the circumference of the body is added, and in the third also the suggestion of a surface. Next to essentialisation, also idealisation plays a role in the creation of the above originals: the small differences in colour of the real blood are intensified to clearly differentiated blue and red. Such idealisation enhances the referring power of the imagesign: it is more evident which vessels belong to the same category and which belong to a different category. Thus, the idealised colours are transformed into a signifying opposition in a colour code. In the imagesign, the construction of an original is not only determined by what is important, but also by criteria of readability as a sign. Something similar can be observed in pictograms that signify male/female: the reduction to black and white highlights the differences in silhouette, which are in their turn reduced to an elementary opposition. The same goes for emoticons.

Such semiotisation leads eventually to the dissolution of the image altogether: exemplary in the transformation of hieroglyphs into the purely arbitrary configurations of the alphabet. Such transformation does not mean that there is a transition from image to sign, but rather that motivation by visual similarity is replaced with absence of motivation when the image evaporates into a purely arbitrary visual configuration.

Let us remark that the construction of a new original has a totally different effect according to whether we are dealing with an image or with an imagesign. Since an image presentifies, we contemplate the original, not the model that served as a starting point in the construction. When we are dealing with a sign, on the other hand, the original, however different from the model in the real world, continues to refer to the model. Even when the original in the imagesign consists only of blood vessels, we continue to relate it to the entire body of which the vascular system is only a part; and even when the vessels in the imagesign are red of blue, we are not searching for vessels of that colour in the real body.

Also the degree of suggestibility or of idealisation/essentialisation, hence, does not affect the status of the sign as such. Below, in analogy to the images of human bodies above, we show six images of human hearts in a similar disposition. If we remove the letters and numbers on the second row, we can read the items either as images of (cut-through) hearts (with cut-off vessels) or as imagesigns referring to the (intact) human heart:

Conversely, all the images in the table of the human body could as well be used as imagesign - were it only for the catalogue of an auction.

The question rises what determines the choice for one of the many different kinds of images. Generally speaking, essentialisation and idealisation enhance the referring power of the image when the obtained elements are integrated into a system of signifying oppositions. Next, criteria of semiotic economy determine which elements in the continuum are chosen - although there is nothing that prevents the use of uneconomic imagesigns: also Dürer's Adam ad Eve would do their job on toilet doors. When we are dealing with simple messages such as those on toilet doors, the degree of essentialisation amounts ideally to leaving only a single meaningful difference, provided recognition remains possible. In that case, strongly schematised pictograms suffice: think of the silhouette of a glass for 'fragile', or of a fork and a knife for 'restaurant'. The more meanings are to be signified, the greater the number of significant oppositions that have to be introduced, and hence the lesser the degree of essentialisation: for referring to blood vessels a such, a line suffices; indicating the difference between two kinds of vessels asks for the introduction of colours; in order to discern more important and less important vessels, size has to be added, and so on; in the limit, only a non-suggestive depiction of a non-essentialised original will do the job.

The conclusion is inescapable also in this second case: the degree of 'schematisation' of the image has nothing to do with the question whether we are dealing with a genuine image or with a sign - be it an imagesign or a symbol. It is not because an image is a true to nature rendering of the original, that it would become a sign (as is often contended in theories of photography: Pierce and the photo as index, Roland Barthes, Roger Scruton). Nor is it because an image is simplified, schematised, or 'abstract', that is would become a sign (Gottfried Boehm). That is not to say that the image would not naturally tend to be 'saturated', because it is meant to be contemplated, whereas signs naturally tend to be as economic as possible - ideally: purely unmotivated configurations - since they are only meant as vehicles for meaning.


Whether we approach the problem form the perspective of the sign - the degree of motivation of the sign - or form the perspective of the image - different degrees of 'schematisation - our inquiry shows that there is an unbridgeable divide between image and sign. The degree of motivation never turns a sign into an image, and the degree of suggestivity, essentialisation or idealisation does not allow to discern image from sign (imagesign or symbol). It makes no sense, hence, to subsume the image under the sign or to construct a continuum from sign to image by constructing a continuum of degrees of motivation or of degrees of schematisation.

Only when we recognise the fundamental difference between image and sign, does it become apparent that they are often functionally combined. Such combination will be the subject of the next section of this text: the entwining of semiosis and mimesis.


Signs are greatly independent from the concrete form they take: 'a' as well as 'a' or 'A' are taken as instances of 'a'. There is no sense, then, in regarding the different appearances of signs as images of a presumed archetype: all the signs are morre or less accurate duplicates. In some cases, signs are genuine serial products - each of them a duplicate, not an imitation of the other: the distribution of a spoken text by radio or television or recording, reproduction of texts in newspapers and books, standard fonts. There may even be deceptive imitations: forgery of handwriting. In the latter case we are dealing with peripheral phenomena.

Things are different when we are dealing with imagined signs, as when we read a text silently and 'hear' the sound of the words. Such imagined words are acoustic images of spoken words. A first form of entwining of mimesis and semiosis, hence, consists in imagining spoken words in our mind. Next to the perceived signs, there are also imagined signs - images of signs. That applies not only to unmotivated signs like words, but also to motivated signs: there is a difference between a perceived and an imagined diagram (or a dito symbol or imagesign).


A second form of the entwining of mimesis and semiosis is to be found in the imagesign: an image that functions as a sign. But there are still other combinations.

Words are normally used as substitute perceptions. But as soon as someone pronounces the words 'once upon a time....', the words are no longer used as substitute perceptions but rather as means of conjuring up the image of what they mean. That is the case in novels and poetry - as is apparent from the fact that we are mostly disappointed when seeing a screen version. To distinguish such signs from ordinary referring signs, we will call them imageconjuring signs. The existence of imageconjuring signs makes it necessary to distinguish immediate mimesis - where the image is directly perceptible - and mediated mimesis - where the image is only imagined.

Since images are mostly considered to be visual, it is not superfluous to remind of the fact that there are images for all the senses: next to exteroceptive sense (ear, eye, nose, tongue....) there are also the interoceptive senses, whose impressions can easily be imagined in the mind. Images conjured up in novels and poetry are mostly quite essentialised - this is so much so, that there is a term for indicating 'saturated' descriptions: ekphrasis (think of Keats' 'Ode to a Grecian urn').

There are more kinds of imageconjuring signs. The written word functions as a normal sign when the text is read aloud: it elicits the reaction of pronouncing the corresponding sounds. But when we read the text silently, we merely imagine the sound of the words pronounced. In that case, the written word functions as an imageconjuring sign. The same goes for a score: the notes elicit either the reaction of performing them on an instrument, or of imagining the corresponding sound in the mind. In the latter case, the notes are imageconjuring signs that conjure up the sound of the corresponding music in the mind.

The most common imageconjuring signs are unmotivated: the words of language. The notes on a score are partly motivated by analogy (pitch). But also symbols can conjure up the image of their meaning, as when the image of the Holy Lamb conjures up the image of the suffering Christ. Also imagesigns or objectsigns can conjure up images in the mind, as when one imagines having bought an item depicted in a catalogue.

It must be emphasised that only exceptionally do signs conjure up images: as a rule, they only elicit an immediate, postponed or suspended reaction. It is an ineradicable prejudice that signs as such do conjure up images ('ideas'). The prejudice stems from the fact that texts are all too easily equated with literature and poetry, and from the fact that language is all too easily equated with single words (Augustine). Thus, it is all too easily assumed that the meaning of 'table' is the mental image of a table - or to phrase it with Wittgenstein: 'Das Aussprechen eines Wortes ist gleichsam ein Anschlagen einer Taste auf dem Vorstellungsklavier' (PU, &6). The way in which words are learned favours such approach. The meaning of the written word 'table' is illustrated by adding an image of the table to the word. In reality, we do not imagine anything at all when hearing the sentence 'There are crumbles on the floor'. We just react by cleaning them up. When reading a text, we only imagine the sound of the words, not the meaning of what we are reading: we just understand the meaning of the text. To be sure, we may try to verify something by imagining some state of affairs - an operation which is sometimes facilitated by adding illustrations. Such verifying, however, is not part of the text itself, but rather interrupts the reading of it. The confusion is enhanced by language. Thus, a metaphor is often understood as a literal image, even when nothing whatsoever is imagined when we understand it. When reading the slogan 'Put a tiger in your tank' we do not imagine that we will buy a tiger, but rather powerful fuel.


Image and signs (with all kinds of motivation, imagesigns and symbols included) are often entwined in more or less complex chains, which we will call 'mimetic chains':

01 config
= sign
02 image
= sign
03 image
04 config
= sign
makes perform config
= sign
05 image
= sign
makes perform config
= sign
06 config
= sign
makes perform image
= sign
07 image
= sign
makes perform image
= sign
08 config
= sign
makes perform image
09 image
= sign
makes perform image
10 config
= sign
conjures up config
= sign
11 image
= sign
conjures up config
= sign
12 config
= sign
conjures up image
= sign
13 config
= sign
conjures up image
14 image
= sign
conjures up image
15 config
= sign
conjures up config
= sign
conjures up image
= sign
16 config
= sign
conjures up config
= sign
conjures up image
17 config
= sign
conjures up image
= sign
conjures up image
18 config
= sign
conjures up config
= sign
conjures up image
= sign
conjures up image

For comments on this table and ample illustrations, see: Entwining of mimesis and semiosis (1): illustrations' (Come back to here by clicking on the link below the attachment).

This table makes it clear that sign and image are often closely entangled, which makes it all the more difficult to distinguish them from one another. From the 18 chains in this table, only 8 end up in a perceived or imagined image: the chains that are coloured in grey. Chain 1 and 4 consist only of unmotivated or analog signs, and chain 10 comprises only an imagined sign (= an image of a sign). Chains 2, 6, 7, 12 and 15 end up in an image that is a sign (imagesign or symbol). They can only be regarded as mimetic chains when these images are not read as signs. And that goes also of the chains that only contain images in the preliminary shackles.

Add to this that from the eight chains that end up in an image, five begin with an unmotivated sign, the sign par excellence What is perceptible in these chains is the sign, whereas the image remains invisible. - exemplary in the book of poetry. That cannot but foster the confusion of image and sign.


To make things even more complicated, all these chains can be combined in their turn. You will find many examples in ' The entwining of mimesis and semiosis (2) (Return to here with the link below the attachment).

Some of these combinations constitute complex images: when there is a combination of chains that end up in perceptible images and chains that end up in mental images (combination of unmediated and mediated mimesis) - exemplary in the song. In other cases (chains that end up in) images are combined with (chains that end up in) signs (images of signs and images that are signs): no pure images, hence, but heterogeneous combinations of image and discourse. In still other cases both chains end up in a sign, whereas only one preliminary shackle is an image - exemplary in the explicit emblem with text.

Let us remark that the combination of mimetic chains must be distinguished from the combination of signs with different degrees of motivation. Let us also remark that, in matters of art, there are countless other forms of mixture: combination of different kind of mimetic mediums and of images for diverse senses. Differences between works of art have to be accounted for with various variables, which are not always clearly distinguished. The sound poems of Marinetti with analog signs for sound volume, are often confused with poems where the form of the words is an image of the meaning (Apollinaire's calligrams). When different senses are addressed, often the same mimetic chain is used: the kernel of theatre and film is formed by mimetic chain (03), where optical and acoustical images are combined. The difference between theatre and opera has to do with still another variable: that of the probability of the original (speaking versus singing humans).

Only through such detailed description of mimetic chains and their combination can we account for the often complicated ways in which image and sign are entangled, without having to resort to mystifying concepts like W.J.T. Mitchells 'imagetext' (where a kind of continuum is postulated between 'imagetexts' dominated by the text, and 'imagetexts' dominated by the image).


Our analysis of the entwining of mimesis and semiosis must be completed by introducing the last and most important shackle - which is at the same time the most problematic.

An original - the object, or by extension: the world - cannot be perceived in its totality: we only can grasp it through countless appearances for diverse senses. From the countless impressions for diverse senses, and from countless verbal communications, we construct a mental 'image' of the world and the objects it contains on the basis of innate or acquired codes of interpretation. That mental image of reality contains far more information that every single concrete sensory impression at a given moment. That is why every concrete perception functions as a sign that activates the entire mental image. Thus, the concrete perception becomes more than what there is literally to be perceived: it is turned into an appearance of an object that is not perceptible in its totality. That goes already for the perception of a 'figure' against a background, where the background is 'perceived' as continuing behind the figure; for the perception of gradients, which are read as rounding and perspective, with the concomitant completion of the foreground with a rear, and of the surface with a content; and so on (see globe below). That goes even more for the interpretation of visual cues in terms of impressions for other senses: the visual appearance of a body betrays how it feels and of what substance it is made. That goes above all for the perception of purely visual configuration on the face, which we read as expressions (= signs) for the inner state of mind of a person (see the Mona Lisa below). And, finally, every perspectival view on a fragment on the visible world is embedded in a complex of spatiotemporal relations (object constancy, cause and effect, history and intentions...): the smoke is related to a fire, the trace with a passing animal, the corpse with the murder (see Marat below):

The perception of reality can be compared with the lecture of a book: we do not so much see letters, as rather understand their meaning. The same applies to the way in which we read configurations of sensory impressions. We see not so much a circle with black stains, as rather an unstained globe with a surface and a content. We see not so much a mere silhouette before a landscape, as rather a body of flesh and blood in the center of a wide space. We see, not so much a mere smile, as rather an inner state of mind in an inner world behind the surface of the face. We see not so much an isolated object, as rather a phase in the unfolding of a complex spatiotemporal whole. Just like with other signs, we read the configurations of sensory impressions as substitute impressions: our reaction is to see their meaning, even when it is not given here and now. That the volume, the soul, the spatiotemporal whole are not given here and now, does not mean that they do not exist: it suffices to touch the globe to feel that it is really there before us and that it is round indeed. But its existence is not here and now perceptible, except then as meaning. To distinguish this kind of signs from other signs, we will call them objectconstituting signs. In that sensory configurations are read as objectconstituting signs, every perception of a sensory configuration is transformed into the perception of the appearance of an object.

The existence of objectconstituting signs is the key to understanding what an image is. It suffices to provide the necessary objectconstituting signs in a mimetic medium that replaces the real world as the bearer of sensory impressions to conjure up the entire object. Just as is the case with the perception of an existing configuration, the configuration in a mimetic medium is read as the appearance of an original. The perception of the image, however, differs from the perception of the object in the real world: when we want to touch the globe with our hands, we stumble on the surface of the canvas - like Narcissus on the surface of the water. That goes also when we continue to look: we cannot see the Mona Lisa in profile. The signified object cannot take other appearances. And that is why we do not talk of a rreal object here, but rather of the imitation: an image.

This approach allows us to define the image in semiotic terms: an image is an original of which the number of appearances is (mostly sensory) reduced, in the limit to one single appearance, like in the non-moving visual image. The concept of the image as a sensory (and/or spatiotemporally) reduced resemblance with an original, is still valid, but it gets its full meaning only against this background.

The signs that constitute an image are still objectconstituting signs, but objectconstituting signs that do not hold their promise: the meaning cannot be signified by a (limitless number of) other signs, because the original has only a limited number of appearances if not a single one: the image. Therefore we prefer to speak of imageconstituting signs. Objectconstituting signs differ also from imageconstituting signs in that many of them are mere images (imitations) themselves: think of the gradient of the globe, the smile on rounded lips which are renderend on a flat surface), in that they are mostly idealised and essentialised.

Also in imageconstituting signs are semiosis and mimesis entwined. This last form of entwining of semiosis and mimesis is problematic for a theory of mimesis. It seems as if the image is not more than a set of signs - and that is what many authors assert indeed. But this contention is nevertheless false: the image is only the meaning of the sensory configuration that functions as an imageconstituting sign. The image is not the circle with a black stain, but a white globe. Asserting that the image is a sign, because, just like the original in the real world, it is constituted by signs, comes down to asserting that a text is not more than a set of signs - the sound of the words or the letters on the pages by which its meaning is signified. The confusion between sign and meaning is enhanced in that in both cases the meaning cannot exist independently from the sign: original as well as image can only exist as appearances. But it is language that seals the assimilation of sign and meaning; just like we use the term 'book' as well for the content as for the pages on which it is printed, just so do we use the term 'image' as well for the appearance of the original, as for the mimetic medium on which that appearance is to be 'read' (the paint on the canvas, and so on). The image in the real sense of the word is the Mona Lisa, not the painted panel on which she appears.

The image is not a sign, hence, but, as we have seen, it can function as a sign - and is then transformed into an imagesign or a symbol: imageconstituting signs conjure up an image of a dove, and the dove can be read as an imagesign (for instance for a dove hotel) or as a symbol (for peace). Because imagesign as well as symbol are mere signs, we often do not await the full constitution of the image. That is why imagesigns tend to become essentialised: the imageconstituting signs that suffice for recognition will do - just think of emoticons. At once, it becomes clear that the semiotic approach of the constitution of an image allows also for a good description of essentialisation and idealisation. Lips can serve for kissing as well as for smiling. When the artist is interested in the smile, the softness of the lips is of no concern, and vice versa.


The above makes it clear how misleading it is to understand reality - the world, the universe - as a kind of visible spectacle of which we could make some image in texts or books - the so called 'worldview'. A 'worldview' is only a view - or an image - in a metaphorical sense, for the simple reason that the world is not a perception, let alone a perception of appearances, but an endless series of perceptions of (found or manmade) signs that serve as the building bricks for the construction of a mental world, only parts of which can be perceived as appearances. Statements about the world - from particular and local, to general - are by no means images of the world, but frames that help us interpret concrete perceptions. Through visualisation, considerable parts of the world can be made perceptible, but visualisations are merely substitute perceptions of a part of the world.

The misconception that the world would be a spectacle that could be mirrored philosophically or scientifically in a book, finds its origin in the fact that we always situate ourselves in an environment - a visual 'world' - which coincides with 'the world' for many animals and primeval man. It is profitable to permanently perceive our environment, also when we are gratified and when there is no actual threat, to be prepared for new opportunities or to escape unforeseen dangers. But how much the visual world wherein we are embedded seems to fill our entire field of vision, it is and remains a perspectival view on a minute fragment of the visual and invisible world. To be sure, we can get a broader perspective until eventually the entire earth comes into view, but not only countless details come to escape us then, but also everything that happens behind us, let alone everything that has happened in the past and that will happen in the future, not to mention everything that is invisible altogether. Reality consists from the beginning of a set of local perceptions of appearances of objects or states of affairs, which are relayed with one another into a trajectory in that they elicit reactions that realise new perceptions. Such trajectory eventually ends up in the elimination of perception as such or in the perception of local pleasure, and such pleasurable perception is always a perspectival view on a rather restricted here and now in an altogether inhospitable reality. We find ourselves not so much in a perceptible world, as rather in a kind of semiosphere disclosed by whatever happens to be in our field of view at a given moment - spoken words and texts included. A 'worldview', then, is only an image in the metaphorical sense, just like a mimetic medium is only a homonym for an image in the literal sense. Conversely, an image is never an encompassing view on 'the world', but always a mere perspectival view on fragments of it. The artist can try to expand the perspective, but it has always to remain rather limited (see: 'The image and its genres', soon on this site). The other side of the coin is, however, that an image is really an image in the full sense of the word: only the image can show us (a part of) the world.

The distinction between a metaphorical 'worldview' and a literal perspectival view on (a minute part of) the world, allows us to understand the difference between art en science/philosophy in terms of the opposition between mimesis and semiosis. Whereas art shows us a world by means of presentifying images, science and philosophy formulate statements about it by means of referring signs. The drama ''Otello' shows us jealousy at work, it is not a treatise on jealousy. Making statements may be facilitated by inserting examples - imagesigns that show how the absent reality referred to in the text looks like. And in making statements, also imagesigns and symbols may be used - think of Plato's parable of the cave. But, when the image is not used to show but to talk about, it is no longer an image, but rather an instrumental image - a sign. It does no longer belong to the world of art, but rather to the world of discourse. With literary allegories like 'Gulliver's Travels' van Swift, Kafka' Das Schloss' or Bertolt Brechts 'Mahagonny', or with pictorial allegories like those of Bosch en Brueghel, the contribution of the image remains convincing and enjoyable in itself, so that we enjoy not only the 'philosophy', but foremost the 'art', although the joy is no longer unbroken like with a Homeros. It is only a pity that the same can hardly be said of Duchamp's 'La mariée mise à nu par ses célibataires même'. Whereas science and philosophy do not show us a world, but make statements about it, the artist does not make statements about the world, but shows us one. To be sure, also the artist's staging of the world is, just like pointing with a finger, a kind of statement: 'ecce'. But this statement amounts to nothing more than saying 'That is how the world looks in my eyes''.

That does not prevent many an artist from having the certain feeling that he expresses himself in what is commonly called 'the language of painting', 'the language of music', or even 'the language of poetry'. In conjuring up a world, the artist resorts to signs, that can only be called a 'language' in the metaphorical sense of the word: imageconstituting signs derived form the objectconstituting signs wherewith we construct our world. Through selection, essentialisation and idealisation, the artist can determine how his world will look like. But, what that world 'expresses' is not the expression of the artist, who merely says 'look here' - and thereby 'gives the word' to a world that can only be staged through mimesis. The so-calledlanguage of painting (or of music), then, turns out to be rather the 'language' of objectconstituting signs. At best, one could say that the artist stages a world to let it say what he wants to say - that the artist is a special kind of ventriloquist: one who does not lend his human language to animal puppets, but rather one that resigns from expressing himself in human language and resorts to the objectconstituting signs that the real world uses from way back. Also what the artist does more than just showing, then, is merely a speaking in the metaphorical sense: the speaking of a world that is staged by him. And that metaphorical speaking should under no circumstances be confused with the use of images as a language in the real sense of the world: the non-verbal speaking with imagesigns or symbols. The 'language' used by the artist is the 'language' of the imageconstituting - worldconstituting - signs.


After the completion of the second part on the entwining of semiosis and mimesis, we can conclude with some general statements.

Man constructs himself a world on the basis of countless perceptions that are interpreted as signs on the basis of innate or acquired interpretation codes. Man enlarges his world by producing new signs, especially verbal language. Thus, he can rely on the experience of others on other places and times. All these signs are means of following a trajectory that leads to the absence of disturbing perceptions or the presence of pleasurable ones. Only the latter are no signs, but ends in themselves. Only when these goals are achieved does the world dawn in the real sense of the word: as a perspectival view from the here and now, mostly for one sense. The image is a means of presentifying such experiences, without having to follow trajectories in the real world, or of reaching goals that simply do not exist, to where no trajectories whatsoever lead.

Although image and sign have to be clearly distinguished, they are working together in many ways. The image is constituted by imageconstituting signs, it can be conjured up in the mind with imageconjuring signs, it can be produced according to the prescriptions of a text or a score, but it is nevertheless no sign itself: it shows a world that has to be enjoyed as an end in itself, no objects that are mere signs. The image can be used as a sign, but it is no longer an image then, but a special kind of sign: an imagesign or a symbol, that is part of the whole array of non-verbal 'languages'. When the image is not used as a sign, it just shows a world. Such showing of a world presentified through the use of imageconstituting signs is different from making statements about the world with verbal or non-verbal signs - from everyday conversations to science and philosophy. Presentifying images are used to show a world, referring signs to make statements about the world.

Showing a world that 'speaks' to us, versus (verbal or non-verbal) making statements about the world: that is the gap that yawns between art and science/philosophy, and it is the same gap that yawns between image and sign, between mimesis and semiosis.

Reasons enough to clearly distinguish the world of the image from the world of the sign - to plead for a genuine 'iconic turn': the development of a theory of the image - of mimesis - that breaks once and for all with the semiotic subsumption of the sign under the image.

Stefan Beyst, May 2010 (translated June 2010).

 facebookshare facebookvolg    twitter

image of the week:

eXTReMe Tracker