Beyond Gombrich: the recrudescence of visual semiotics

Da: Journal of Art Historiography, Number 5, December 2011.

Point at Issue

Born in Vienna to a bourgeois Jewish family, Ernst Gombrich grew to be a scholar familiar with transitions. Having moved from his home country to settle in London in 1936, Ernst Gombrich soon felt the need to translate his own culture to another system of knowledge. In the passage from Viennese to Anglo-Saxon culture – as in all translations – something was lost, but much more was gained. And, as every life begins incomplete, yet full of potential, so too does the body of a great intellectual’s work. It is now time to realise the potential of Gombrich’s academic career and also to define his semiological roots.
From the very beginning of his career, Gombrich made use of Karl Bühler’s reflections on sematology, which was the main frame of reference for the semiotics of those days. The aim of this text is to place Gombrich’ work (on the centennial anniversary of his birth, and hence this phase of revisionism), within a framework of research. In the deixis of hic et nunc, this text will be anaphoric – referring to early Gombrich and his relationship to Popper, Bühler and Panofsky, and cataphoric; referring to what will follow. After all, it was Bühler who introduced the notion of ‘cataphora’, as reference not to the past, but rather orientations towards the future. An open cataphora is, in my mind, a truly auspicious sign for a conclusion of this venture.

1. When one is not an art historian

First of all, let us ask: is Gombrich an art historian? In the opinion of James Elkins he is not, and there are ten fundamental reasons that explain why1. According to other sources, Gombrich has scarce interest in the history of contemporary art, considering it eccentric or even rejecting it2.
Elkins’s argumentative strategy is to list one-by-one generalities and their connotations within art history. Elkins enumerates these points and then accurately demonstrates that none of them fall inside Gombrich’s Weltanschauung. The first is the relationship to progress, to evolution. He who believes in it, Elkins points out, can be considered a historian (of art). The second is the process of periodization, i.e. accepting certain kinds of divisions and admitting, for example, that mannerism was a period, which was followed by the baroque, rococo and so on. The third is the supposition that art is a reflection of the mentality of an era, a spirit of the times or Zeitgeist. The fourth is conforming to write the history of art primarily in the form of monographs about individual artists or illustrating periods. The fifth is practicing attributionism.
According to Elkins, if these are basic parameters to defining the profession of art historian, then it is easy to see that Gombrich does not fall within the profession. Indeed, he would appear to be a dedicated anti-historicist insofar as he is anti-Hegelian and anti-evolutionist, and instead embraces Popper’s idea that nothing can be proved if not through the falsification of concepts. Furthermore, he is convinced that progress and evolution are limited to specific relations and specific ends, i.e. that they function within a particular system of representation or techniques. He does not write authorial monographs and as a rule he does not make attributions. When he does, it is to show that attributions are farcical and that they use risible assignment criteria. Many will remember the case, constructed as an example of absurdity, of the prophets in the Sistine Chapel: Gombrich attributed Michelangelo’s Ezekiel and Jeremiah, with an ironic abundance of proof (though still dangerously insufficient), to Raphael, if not to Leonardo3.
Wishing to apply Popper’s method of falsification, it comes to us to show, on the other hand, which are the pertinent traits that attest to Gombrich’s place within the canon of art history. Several problems arise here. For example, although Gombrich believes in periodization, his interpretation of the concept does not conform to the canon. For him the avant-garde begins with the renaissance, time of the first artistic commissioning and competition among artists, time of the first records4. Then Gombrich distinguishes between periods and movements5. Thus, mannerism, which is metalinguistically labelled from the outside, is a period; while futurism or the renaissance, assumed and practiced by the artists themselves, are movements. They should not be placed on the same plane, nor should they be treated historically in the same way.
There is, however, another solution, which is to argue that Gombrich is a theorist of communication. Naturally this is a delicate thesis because, in this particular case, we cannot limit ourselves to the cyclical relationship of sender-receiver-message. Gombrich does not treat art like all other messages; rather he places it within the framework of other types of messages. Perhaps it would be more appropriate to ask whether he is a theorist of artistic communication, and enquire how artistic communication relates to other types of communication. In this way, a Goodmanian6 approach, Gombrich could help us to understand not if, but when a form of communication is artistic.

2. Gombrich and sematology

Let us now return to sematology. To understand Gombrich’s modus operandi it is particularly important to examine the difference between sign and symbol. Indeed, Gombrich himself noted: ‘all images are signs, and the discipline that must investigate them is not the psychology of perception—as I had believed — but semiotics, the science of signs’7.
A student of Bühler, who was a sematologist, Gombrich worked his entire life within this paradigm. He loved recounting the episode of a lady who, standing before a portrait of her friend painted by Matisse exclaimed: ‘one of the arms is too long; my friend’s arm isn’t like that’. To which the artist replied: ‘but Madame, this is not a woman – it is a painting’. Yet Gombrich also loved the following affirmation by Daniel-Henri Kahnweiler: ‘la peinture ne se justifie que si on la considère comme une création de signe et non d’objets feints‘ (painting is only justified if it is considered as a sign and not a re-creation of feigned objects)8. Here, the theoretical implications regarding perception are enormous, because as James Gibson rightly explains, we do not perceive external reality in the same way as we perceive the brush marks on a canvas, which have special rules that allow us to recognize (or not) certain figures. The perception of a sign is not the same as the perception of an object existing in the real world. From this point of view, Gombrich agrees with him.

2.1. The relationship between signs

Karl Bühler, certain that the objects of perception were signs, i.e. expressive forms connected to forms of content and with an active and transformative power, elaborated a typology of images: diagrams, coats-of-arms, scientific instructions, photographs, postage stamps, portraits in wax, statues and paintings. Thus he puts on the level of visual study an array of signifying elements with different values compared to the traditionally accepted elements and opens the question of mimesis. Cesare Ripa, in 1593, represented ‘painting’ as a gentlewoman intending to imitate something, but with a strange tool in her hand, a sort of ruler, a symbol of correction. Mimesis is not imitating things but correcting them, in our way of reproducing them and in their own nature. Before reproducing something, it is better to correct it, particularly in function of its representation. Thus maps, family trees, Darwinian rhizomatic representations, coats of arms etc. finally belong to the realm of discourse. No longer are they images excluded from the artistic system due to differing ends; procedural rather than aesthetic. They exist not only for their likeness to something real, but for their likeness in relations, which are in part recovered, in part discovered, in part created.
In a 1972 article in Scientific American, Gombrich accurately described a complex image of great importance which is now voyaging outside our solar system and for several decades has been moving in the direction of some imaginary planet9. It is the plaque travelling aboard the space probe Pioneer 10, a sign made of aluminum, conceived and designed by NASA engineers to be intercepted by extraterrestrials (Fig. 1). Its aim is to transmit information about ourselves, as human beings, to someone that we have not previously met. How does one go about designing a message for a virtual recipient? The image at the top left, corresponding to the arrow, represents the hyperfine transition of neutral hydrogen, which is the most abundant element in the universe. Nearby there is a vertical line which indicates the binary digit one. The ‘spin flip transition’ of a hydrogen atom from its state of electron ‘spin up’ to the state of electron ‘spin down’ can specify a unit of wavelength (21 centimeters) and a unit of frequency (1420 megahertz).

Pictorial plaque on the 'Pioneer' F Spacecraft. 1972

The geometric shapes immediately behind the male figure represent a diagram of the Pioneer, but on a different scale from the people and depicted in a top view (thus from a different perspective), to show there is correspondence between the two elements. How many conventions are needed to understand it! That the man with raised arm is standing in front of the Pioneer is absolutely a convention of our culture. In the text there is nothing to show that in actuality the man is not completed by a beautiful geometric halo, which is not behind but around him. At a certain point the author of the image must have abandoned the idea that the two humans could be holding hands, which is a reasonable solution. There was the risk that the recipient may think of man as a being habitually conjoined to another part of himself. Similar devices were not employed to contrast the impression that on our planet some beings live with their arms lowered (women), while others (men) move around with one arm raised and the other lowered. Similarly, the representation of the human race as white and occidental is ambiguously racist. It seems commonplace to show oneself nude in public, though evidently a sense of decency at NASA advised suppressing the female genitals, conserving only the pubic mound. The man appears with two similar hands although asymmetrical; the woman, for some unknown reason, has one ‘normal’ hand while the other looks somewhat like an octopus wrapped around her right thigh.
In the lower portion of the picture a series of lines is intelligible, at least to us, as the center of our galaxy, with the fourteen largest pulsars of our solar system. Beneath that we can see Earth, the sun, and our neighboring planets, each at the correct distance from the sun. They are connected to a line that, leaving from the third planet, suddenly skips around the next two, passing before another one, Saturn, which is instead in front of the woman. The line ends in an arrow which points to yet another figure, the Pioneer. An iconographic problem of extreme interest emerges. We must assume that the recipient culture has at some point practiced hunting or warfare with arrows. We cannot take for granted that they are familiar with arrows or understand their symbolic meaning. Finally it is unclear why the arrow indicating the Pioneer is of dimensions similar to Saturn’s rings.
The entire image presents signs so different in nature and conformation that it is impossible to speak about immediate perception. On the contrary, the message implies an elaborate and abstract quantity of knowledge and an incredible number of conventions such that the target recipient, the extra-terrestrial, may not understand or be interested. In many instances they are symbols. And here we need to clarify the difference between a sign and a symbol.

2.2. The ritual aspect of symbols

In the rereading of Warburg, one has the impression that the concept of a ‘symbol’ has remained the same since being outlined in the nineteenth century by Hegel. Symbols seem to be something mysterious, complicated and with religious connotations; on the opposite side there are codes with only one meaning10. Language, which is made of signs, would then create a field in which every word has precisely one meaning. Poetry would have to render through semiotics each word’s assignation to a single symbol, typical of linguistic codification, while symbolics would be the place to examine the mystery and complexity of life.
Edgar Wind’s violent attack in 1971 against Gombrich’s study on Aby Warburg comes back to mind. For Wind, Warburg had masterfully understood the functioning of symbols, elements in which signifiers and signified are fused together by ritual aspects11. In between there is art, which even while using signs, would be able to maintain an intensely ritual dimension of life. Here, Wind summons the theory of Friedrich Theodor Vischer12, first accepted within the field of art history and later to be elaborated upon in philosophy. Vischer, who was a correspondent of Croce, defined the symbol as a connection between image and meaning, through a point of comparison. What was fundamentally important to him were not the types of signs, but the types of connection between image and meaning, which take into account the participative nature of the subject. He thus distinguishes between a magical-associative connection and a logical-dissociative connection. By now these ideas have been greatly surpassed. It is sufficient to mention Karl Popper, who resuscitated Bühler’s ideas with his own functions – expressive (Ausdruck, of the sender’s feelings), descriptive (Darstellung, of states of affairs), and argumentative (Appell, focused on the receiver); to understand there was a break.

3. The outlook for artistic motifs

From the point of view of a ‘semiotic recrudescence’ it is possible to further highlight Bühler’s influence on Gombrich through two paths: the first improves a historicist and symbolist reversion; the second endorses a discursive, textual and enunciative advancement.
Gombrich wanted to construct a linguistics-based theory of image founded on a Darwinian-based evolution of styles. In his view, the research of Alois Riegl had moved in the direction of an etymology of ornamental patterns13. Of the Caucasian rugs in Riegl’s collection, he affirmed: ‘these visual poems have a Greek accent, later Egyptian’. When Bühler considered how Charles Darwin observed expressive gestuality, he had also declared himself favorable to the introduction of certain contrastive criteria. Darwin had already offered relevant observations, with the theory of the antithesis, but in general they were reflections that derived from his interest in contemporary phonology. This is the starting point from which Gombrich develops the idea of a linguistics of the image, that is, a visual semiotics.
In his review of Morris’s volume, he writes:

The distinction between poetry and language has always been accepted as natural; the distinction between art and imagery is only gradually becoming familiar. Mr. Morris himself stresses the need for more descriptive studies on visual signs. It is all the more a pity that he does not seem to have taken cognizance of the emerging discipline of iconology, which must ultimately do for the image what linguistics has done for the word […]. The works of such pioneers in the study of the symbolic aspect of the image as A. Warburg and E. Panofsky are absent. Iconology was not the study of complex emblems and allegories but the interaction of forms and contents in the clash of traditions14.
At present there is a widespread and negative Warburgian rereading of Gombrich underway. This requires careful observation. Warburg also engaged linguistics; notably the nineteenth-century historical grammar of Hermann Osthoff, prior to both the turning point with Saussure and the Prague Linguistic Circle. For Warburg the Pathosformel had a semantic and formal value, alongside an irrefutable historicist character. Pathetic configurations function as binarily opposed motifs. Warburg had a particular interest in gestures characterized by excessive pathos, ‘superlative’ (from a linguistic perspective), and Dionysian, as opposed to Apollonian. We should note that Pathosformeln are not separate but are described in couples.
The binary evolution of pathemic forms brings Warburg closer to linguistics and structuralist semiotics. It is therefore a pity that these fundamental motifs, inscribed in the pathemic structuring of man himself, have been conceived and used only as documentary sources to reprise in modern and contemporary texts. Semiotics does the opposite: It marries the existence of figurative motifs endowed with meaning and capable of moving between different cultures and texts, (even within a single text), but it also accounts for the changes in form and meaning which the motif undergoes, when it is inscribed within a new culture or a new text.
The choice is therefore between a lexic of motifs, to be examined in their emergence and their transformations, or a lexic as invariant and texts as variables in which the motif is inscribed. Is it preferable to study a lexic to find in all texts a meaning that is pre-established a priori, or is it better to see how these motifs inscribing themselves in contemporary texts assume meanings that are completely different? In my opinion the strategy employed both by Gombrich and the latest generation of semiotics is of the second type; it intends to demonstrate that visual texts, whose organization is invariant, gather motifs and transform them in differentiated variations. On the contrary, if the study of Pathosformeln is resolved with the verification of ‘loci‘ (in which to rediscover configurations of the past), our option will be historicist.
On this problem, Gombrich was not sufficiently clear. He investigated the inscription of Pathosformeln in contemporary art and identified, typologically, paths of recovery by way of imitation (pure reproduction), as per assimilation, or exclusion. Some texts could have recourse to the vocabulary of the motifs, but decide not to do so; other texts do make recourse but sacrifice certain parts; still others explicitly imitate them, which is the opposite of exclusion; and finally others assimilate them with a selective assortment of elements. This is the concept of ‘abstractive relevance‘: in its entrenchment, the motif loses some properties while maintaining others which are essential to constructing a meaning.

4. Deixis in images

One last point. Bühler had proposed a theory of deictic indication, with a vision of demonstrative systems that was quite original, and yet remained hidden for many years. In fact, Gombrich utilized the notion of deixis only in a generic way. This is one of the limits of his approach.
The most productive aspect of Bühler’s theory, the invention of enunciation, was removed in the first place by logisticians, who considered pronouns a bizarre tool for indication. Indeed, the pronoun is very strange, because it never indicates a specific thing but rather something in the perspective of the person speaking in that moment. The second removal came from referentialist linguists following the example of the logisticians. It is problematic for a reference theory to accept the fact that when I say ‘I’, I say ‘I’, and you are ‘you’, but when you speak, you are ‘I’ and I become ‘you’; a complex administration of the presence of reference, which is, however, inscribed in language.
The great communication specialist Émile Benveniste compared enunciation to ‘cette tête de Méduse qui est toujours là, au centre de la langue, fascinant ceux qui la contemplent‘ (‘a Medusa’s head, which is always there, at the centre of language, mesmerizing those who gaze on it’)15. This device, though not taken very seriously, is present not only in language but in all communication systems, including visual communication. Thus Meyer Schapiro who in 1973 worked on the I/you shifting both directly and transversely, gave an extraordinary impulse to Gombrich’s research16. He gave weight to the dialogic in interaction and not only to the iconological quality of the character.
In my view the use of pronouns in communicative relations is central to the study of signification processes, an aspect of Bühler’s academic heritage that was interrupted. It is peculiar that Benveniste never cites Bühler; on the other hand Bühler perhaps was not even aware of Benveniste. But both of them, like Saussure, became established thanks to Indo-European studies. Karl Brugmann, the great linguist and student of Humboldt, is the first to have brought into focus the problem of I/you pronoun relations within Indo-European systems, opening the way for Edward Sapir and others who would later identify the same problem in non-Indo-European cultures. From here a visual semiotics of text and discourse developed. This expanded the investigation of the formal and figurative dimensions and articulated the analysis of systems of viewing17.
If we do not want to return to a historicist analysis of the motif, no matter if binarily organized, and if we want to go beyond Gombrich’s approach, which did not take full advantage of Bühler’s tradition of a theory of indication, then semiotics can offer some advantages. Indeed, it manages to study discursive complexity which involves iconological recognition, narrative and argumentative organization, as well as the structuring of points of view in communicative relations. The discipline encounters two types of resistance. The first is historicist, a perspective which Gombrich actually embraced; and the second comes from the theory of symbols. If current iconology is not able to distinguish between symbols of referential assignation and symbols of interactive and deictic strategies, then its contribution has little worth. Semiotics can break through the impasse and hence become even more productive than Gombrich’s paradigm.
In any case Gombrich’s appeal to simplicity18 is not a refusal of theory; it is a product of his acculturation within the Anglo-Saxon system. It is a way to hide the complexity of the discourse behind the basic English rhetoric that in essence the world is simple and we are prepared to understand it; the latin thesis on the other hand is that the world is complicated and we too are complicated. The ideal would be to demonstrate that the simplest texts are fruit of the maximum quantity of mediations, articulations and connections. It is not the object that is simple, but rather our disposition towards it: the capacity to maintain the proper level of simplification or complication needed to make it intelligible.


  1. James Elkins, Ten Reasons Why E. H. Gombrich Is Not Connected to Art History, The Gombrich Archive, 2005, torna al rimando a questa nota
  2. Arthur Danto, After the End of Art, Princeton, Princeton University Press, 1997. torna al rimando a questa nota
  3. Ernst Gombrich Ways of Worldmaking, Indianapolis and Cambridge, Hackett Publishing Company, 1978. torna al rimando a questa nota
  4. Gombrich, ‘The Logic of Vanity Fair. Alternatives to Historicism in the study of Fashions, Style and Taste’, The Philosophy of Karl Popper Illinois, Open Court, (1974) 1979, 925-957. torna al rimando a questa nota
  5. Gombrich ‘The Renaissance-Period or Movement?’ Background to the English Renaissance: Introductory Lectures, J.B. Trapp, London, Gray-Mills Publishing 1974, 9-30. torna al rimando a questa nota
  6. Nelson Goodman has the merit of liberating us from the ontological question, ‘What is art?’ moving the reflection towards the more empirical and verifiable question, ‘When is art?’ See Goodman Arte in teoria arte in azione, Milan P. Fabbri ed. 2010. torna al rimando a questa nota
  7. Gombrich Art and Illusion. A Study in the Psychology of Pictorial Representation, Oxford, Phaidon 1959 ed. 2000 xxv. torna al rimando a questa nota
  8. Gombrich Art and Illusion. See also Gombrich ‘Voir la Nature, Voir les Peintres’ Les Cahiers du Musée national d’art moderne, vol.24, été 1988, 21-43. torna al rimando a questa nota
  9. Gombrich ‘The Visual Image’, Scientific American, September 1972 and reprinted in The Image and the Eye: Further Studies in the Psychology of Pictorial Representation, London, Phaidon, 1982, 137–61. torna al rimando a questa nota
  10. For an ample and profound explanation of the theories of the symbol, see Eco 1984. torna al rimando a questa nota
  11. Edgar Wind Review of Aby Warburg. An Intellectual Biography, The Times Literary Supplement, 25 June, 1971, 735-736 reprinted in The Eloquence of Symbols. Studies in Humanist Art, Oxford, Clarendon Press, 1983. torna al rimando a questa nota
  12. Friedrich Theodor Vischer Das Symbol, in Philosophische Aufsätze as dedicated to Eduard Zeller on his fiftieth anniversary in his position as doctor, Leipzig 1887. torna al rimando a questa nota
  13. Gombrich The Sense of Order. A Study in the Psychology of Decorative Art, London, Phaidon, 1979. torna al rimando a questa nota
  14. Gombrich Reflections on the History of Art. Views and Reviews, Oxford Phaidon 1987, 246-249. torna al rimando a questa nota
  15. Émile Benveniste Problèmes de linguistique generale, Paris, Gallimard, ed. 1966, 126. torna al rimando a questa nota
  16. Meyer Schapiro Words and Pictures The Hague, Mouton 1973. torna al rimando a questa nota
  17. Algirdas Julien Greimas “Sémiotique figurative et sémiotique plastique”, Actes Sémiotique. Documents, 1984 60, also M Corrain and L Valenti Leggere l’opera d’arte Bologna Esculapio 1991. torna al rimando a questa nota
  18. Gombrich Acceptance Speech for the Seventh International Congress of Germanic Studies (Göttingen 1985) as printed in ‘Topics of our Time’, London Phaidon, 1991. torna al rimando a questa nota
Print Friendly, PDF & Email

Lascia un commento