The reputation of the Milanese painter Giuseppe Arcimboldo is based entirely on a dozen or so bizarre pictures showing portrait-like heads made up of animals, plants, or inanimate objects. The most famous are the four Elements and the four Seasons. Fire, for example, is a combination of burning coals, guns, fuses, and the like; Air is formed of a multitude of different birds, Water of fishes, and Earth of quadrupeds. The Seasons, by contrast, show flowers, vegetables, and trees. A portrait of a librarian is composed of books; another, of a well-known jurist ravaged by syphilis, has a body made from legal documents and a head formed of plucked fowls and a fish. The Cook and The Vegetable Gardener are based on a slightly different conceit. When displayed one way up these images look like human heads; but when reversed they show respectively a platter of roasted meats and a bowl of vegetables.

From such little evidence as survives it seems that the early work of Arcimboldo, who was born in 1527, was entirely conventional. From 1549 he is known to have been employed at the cathedral of Milan, mainly in the production of designs for stained-glass windows, and in 1558 he made drawings for tapestries for the cathedral of Como. In 1562 he entered the service of the Austrian Hapsburgs as court portraitist, and continued to work for them until he died in 1593, six years after his return to Milan. The earliest securely dated examples of the bizarre heads are the Seasons, of 1563. The last, a portrait of Rudolf II in the guise of Vertumnus, composed principally of fruit and vegetables, was sent to the emperor from Milan in 1591. During the period he spent at the imperial court Arcimboldo produced several versions of the composite heads, including at least four separate sets of the Seasons, but he seems to have been mainly occupied in providing designs for court festivals such as pageants and tournaments, for which a number of his drawings still survive. Very little is known of his activity as a portraitist, but the few traces that exist are unremarkable.

“The Arcimboldo Effect,” held last spring at the Palazzo Grassi in Venice, was the first exhibition ever devoted to Arcimboldo. It was divided into two main sections. On the first floor was an impressive group of the composite heads, represented mostly by originals of high quality but in a few cases by copies of lost pictures, thus fully illustrating his output in this genre. There was also a generous selection of his drawings, together with paintings and prints by contemporary and later imitators. The upper floor was given over to a more miscellaneous selection of twentieth-century works, with the Surrealists and Dadaists most strongly represented. These were included not so much to illustrate the direct influence of Arcimboldo in the early part of this century, which seems to have been very slight, as for considerations of a more ambitious, if also much vaguer, kind. As Yasha David, the exhibition commissioner, explains in the foreword to the catalog:

The comments of his contemporaries on the symbolical meaning of Arcimboldo’s work and the repetitive interpretations of today do not address the basic issue of the real significance of his gesture in intervening on the human face to reconstruct it. It is clear that humanity is being put to the question, thereby opening a new aspect in our vision of the multi-dimensional man.

Despite David’s apparent lack of interest in the problem, it was evident that some of his colleagues were more concerned to explore the significance given to Arcimboldo’s work in his own day. This would be a difficult thing to do in any exhibition. But simultaneously to set Arcimboldo within his own cultural context and to treat him as a forerunner of a type of sensibility that fully flowered only in the twentieth century was to attempt altogether too much. The result was an exhibition incoherent in its organization and superficial in its presentation of ideas.

Each room had a different theme. Some, such as “Arcimboldo in Milan,” were quite specific, while others were vague and portentous, for example “Perspective of the face,” “The testimony of the eye,” “Break in ideas.” Since a similar sequence of these more generalized titles was given to rooms on both floors, evidently the intention was to suggest that Arcimboldo’s innovations in the representation of the human head, like those of twentieth-century art, were in some way related to a wider challenge to conventional ideas. Just how seriously this was meant to be taken, however, is not at all clear, since the objects displayed were inadequate to illustrate the subjects indicated by the titles of the rooms and the issues themselves were discussed only tangentially in the catalog.

Thus “Break in ideas” on the first floor was illustrated by a couple of small cabinets containing a few books, such as the illustrated volume on human anatomy by Valverde (1560), Alciati’s collection of emblems (1531), and Fracastoro’s poem on syphilis (1530). For the modern period there was an unsurprising selection of works by Husserl, Freud, Einstein, Nies Bohr, Kafka, Nietzsche, and Apollinaire. Evidently both collections of books were intended to convey the sense that new ideas were in the air. What those ideas might have been and what possible relevance they might have had to the other exhibits was never explained. Perhaps the most useless of all the captions was “Magic Prague,” which served as the label to two display cases, each of which contained a view of Prague made of colored stones, a medal of the Emperor Rudolf II, and a scientific instrument. Elsewhere in this room there were a few curious objects and carvings, including a painted horoscope and a mandrake root, and a pile of books, most of them closed.


Most visitors must soon have realized that this was not meant to be a particularly didactic exhibition. Like window dressings, many of the objects displayed were simply there to puzzle or entertain, or both. Thus two rooms on the first floor were given over to a selection of anamorphoses and other perspectively distorted images of the human head; but only a very attentive reader of the catalog would have come upon a footnote which implied (probably incorrectly) that Arcimboldo himself had produced works of this type, which no longer survive. The general intention was to make the works of art more interesting by hinting at a connection with difficult or esoteric ideas, without attempting to explain the nature of those ideas or their relationship to the objects themselves: in short, an approach of deliberate mystification and cultural name-dropping, which made the presence of the Surrealists and Dadaists particularly appropriate.

The upper floor was also the most attractive and varied part of the exhibition, and included a number of works of outstanding quality. It may have been a more considerable achievement on the part of the organizers to have succeeded in bringing together most of Arcimboldo’s authentic paintings and related prints, together with a number of early copies, including no fewer than four versions of The Librarian. But few visitors could have been convinced by the claim of Pontus Hulten that “the quiet beauty of [Arcimboldo’s] panels, a serene musical melancholy, is similar in tone to Vermeer’s work.” Arcimboldo had one good idea, which he exploited with considerable ingenuity. His followers and imitators, unfortunately, proved for the most part to be altogether less witty and inventive.

The frivolousness of the exhibition was most clearly demonstrated by the catalog, now published as the book under review, which in fact is not a proper catalog at all, but a collection of twenty-one essays by various authors. It is divided into two parts, roughly corresponding to the two sections of the exhibition. The first part contains a series of articles on aspects of Arcimboldo’s work and on Hapsburg culture. Several of these are informative and useful, notably Piero Falchetta’s anthology of early criticism about the artist. But in general the authors are concerned to address specific historical problems or to advance their own theories about particular pictures. What is missing is any clear general account of Arcimboldo’s career, of his place in sixteenth-century art, and of the current state of knowledge about his output. The second section is even less coherent. Falchetta provides an anthology of modern criticism, of which the most striking is a brilliant attempt by Roland Barthes to interpret the composite heads in terms derived from rhetoric. But too many of the other contributions consist of self-indulgent theorizing that often seems remote from the concerns of artists in this century or any other. In short, the catalog is like the exhibition itself. It is good to look at, full of attractive illustrations, but lacking any clear focus, organization, or purpose. As one might expect, none of the illustrations is numbered and many of the objects reproduced are not discussed in the text. As a result the volume was useless as a guide to the exhibition, and this was all the more regrettable because the labels beside the exhibits were often inadequate.

The clearest impression that the attentive reader would obtain from the catalog is that scholars now believe that far from being mere curiosities, visual jokes, Arcimboldo’s paintings are complex allegories of monarchy, drawing in some ill-defined way on the more esoteric ideas current in the later Renaissance, such as hermeticism, hieroglyphics, the Art of Memory, and Neoplatonism. In addition to the widespread scholarly belief that almost all Renaissance paintings are in some way didactic or edifying, there are two main reasons for this view. The first is the fact that they were painted for the Austrian Hapsburgs, and that one of Arcimboldo’s patrons, Rudolph II, was certainly an eccentric man drawn to the less conventional currents of thought of his day. The second is that two of Arcimboldo’s contemporaries provided very elaborate interpretations of some of his pictures. One of these was a Milanese writer named Giovanni Battista Fontana, or Fonteo, who spent some time at the court of Maximilian II, while the other was Gregorio Comanini, from Mantua, who published a dialogue on poetic and pictorial imitation in 1591.


Fonteo collaborated with Arcimboldo in at least one court festival, in 1571, but it seems that relations between the two men were not especially easy, because twenty years later the painter recalled that on this occasion Fonteo had tried to take credit for his inventions. At the beginning of 1569 Fonteo presented Maximilian with a poem about the Elements and the Seasons. It has been suggested that these pictures were given to the emperor at the same time, and that Fonteo’s poem was written to explain the significance of this gift. But since the pictures had been painted some years earlier this proposal is not entirely convincing. Be that as it may, the poem is a long and convoluted panegyric of the emperor, intended as it were to give “a voice to the royal pictures, and in the hope that the work will be more welcome.” The text has never been published in full, but the basic theme seems to be that the harmony of the elements and seasons wrought by Arcimboldo symbolizes the power and clemency of the emperor and proclaims that his family is destined to rule the world in future ages.

This is all rather clever, but it is the kind of cleverness that would be expected of a court poet, not a court painter. There is no reason, in fact, to suppose that Fonteo’s ingenious conceits are anything other than ex post facto embroidery on Arcimboldo’s pictures. Exactly the same kind of panegyric was produced a few years earlier by Giorgio Vasari on the basis of a series of mythological frescoes that he himself had painted in the Palazzo Vecchio in Florence. Here almost every myth was given an allegorical meaning, often of a very far-fetched kind, relating the subject to the patron, the ruler of Tuscany. But it has recently been demonstrated that these topical meanings were devised after the pictures had been painted.* This is what also seems to have happened with Arcimboldo’s paintings. The artist had shown his wit by representing the entirely familiar themes of the elements and the seasons, which are among the staples of sixteenth-century secular art, in a new and fanciful way. Fonteo, in turn, displayed his own powers of invention by weaving a web of flattery out of them.

A couple of decades later Gregorio Comanini produced an entirely different but equally elaborate exposition of the Elements. For example, he gave a specific significance to many of the animals that make up the features of Earth, suggesting, in one case, that the elephant was chosen to represent the cheek, which exhibits shame, because this animal has a highly developed sense of shame and only copulates in privacy. It is all wonderfully ingenious, but, just like Vasari’s exposition of his own frescoes, it only accounts for some of the details. The rest are passed over in silence.

Although it is unlikely that he would have been able to understand Fonteo’s convoluted Latin, Arcimboldo might well have enjoyed such displays of literary wit. He certainly cannot have been surprised by them. After all, because his pictures are so full of entertaining details and because they do not come from the repertoire of subjects that were usually given standard interpretations, such as exemplary episodes from ancient history, they were particularly amenable to the kind of fanciful exegesis that literary people, trained in the interpretation of emblems and personal devices, were capable of providing for the most unpromising and prosaic material. But the meanings that were applied to Arcimboldo’s paintings varied according to the circumstances in which they were devised. Because Fonteo’s poem was addressed to the emperor, it was inevitable that he should have explained the Elements as imperial allegories, whereas Comanini only provided this kind of reading for Vertumnus; and in this instance he could hardly have done otherwise, given that it was a portrait of Rudolf II. But he certainly did not try to give complex interpretations for all the paintings by Arcimboldo that he described. For example, he praised The Jurist simply as an amusing joke.

If the interpretations of Fonteo or Comanini tell us something about the way in which some of Arcimboldo’s paintings could have been discussed under certain circumstances during the sixteenth century, they do not explain why these pictures were painted in the first place, nor what they might have meant to Arcimboldo himself. To solve these problems we need to consider both the context in which they were produced and the kind of knowledge that artists were then expected to bring to the production of works of art.

In the few contemporary texts in which Arcimboldo’s compositions were mentioned they were almost always called grilli, capricci, or ghiribizzi. These were terms commonly applied to grotesques—decorative paintings containing details such as swags, candelabra, and little figures and animals often combined into bizarre monsters, arranged in a wholly nonnaturalistic way. Such compositions were greatly admired in sixteenth-century Italy. Not only were they sanctioned by antique precedent, because examples had been found in buried Roman buildings (grotte), hence the name grotesques, they were also thought to provide an opportunity for artists to display powers of invention in a way not available to writers. This was a matter of some importance to painters and sculptors.

During the Renaissance the activity of artists was regularly assessed in similar terms to that of poets, because both were concerned with imitation. But whereas poets imitated emotions and ideas as well as figures and landscapes, artists were limited to the imitation of the physical world. The poets drew on philosophy and all kinds of speculative knowledge; the painters dealt only with sense-impressions, and so their activity was to some extent intellectually inferior. Of course, painters and sculptors were expected to improve on nature by a process of idealization, but this was also true of poets. In the creation of grotesques artists came into their own, because this type of art demanded a specifically visual inventiveness, involving as it did the depiction of creatures that had never existed and that were entirely a product of the imagination. They were like images seen in dreams, and they called for qualities not exercised by poets; for Horace had written of such paintings at the beginning of his Ars poetica, saying that they provoked laughter, and had then gone on to condemn the use of similar incongruous conjunctions in poetry.

Sixteenth-century art critics provided elaborate criteria for assessing the quality of grotesque decoration. In particular, artists were expected to observe decorum and a degree of plausibility, maintaining the correct proportions of the component parts and combining them in a manner that was elegant and subjectively credible. But grotesques were generally regarded as being almost devoid of meaning. Even the Milanese artist Giovanni Paolo Lomazzo, the one writer on art of the later sixteenth century who was drawn to esoteric subjects such as magic, concurred with this view. He conceded that in antiquity such decorations may originally have functioned as hieroglyphs, concealing hidden significance, but he did not recommend that they should be treated in such a way by modern artists. Instead, he suggested that their content should be appropriate to figurative paintings nearby. Altars and sacrificial vessels, for example, would be suitable components for grotesque decoration near pictures of religious rites, and figures of Fame would be suitable near representations of famous men.

Arcimboldo’s bizarre heads fit very well into this complex of ideas. Indeed, they are among its most spectacular manifestations. Not only could the spectator admire the skill with which each component was depicted, he could also marvel at the imaginative way in which these were combined to create something entirely different, a head, and one moreover in which each part was perfectly appropriate, thus fulfilling the admired quality of decorum. Winter is not merely composed of wintery plants and trees: he has the physiognomy of a rough old man; while Spring is a youth, and each successive season is older than the preceding one. The power of Arcimboldo’s invention was thus displayed in a way entirely fitting to a painter, in the combination and manipulation of things, not ideas. His pictures even conformed to the characterization of this kind of art provided by Horace, for they were thought to be amusing. Lomazzo suggested that they would provide suitable decoration in taverns: and, as Comanini observed, they were even more fanciful than images in dreams.

For some reason the contributors to the catalog seem to regard this kind of interpretation as inadequate. Even though they are prepared to concede that these pictures are jokes, they insist that “they are serious jokes,” whatever that may mean. They want Arcimboldo to be clever, but in the manner of a writer, not a painter. It comes as no surprise that the catalog includes much more about the cultural ambience of the eccentric Rudolph II, for whom Arcimboldo certainly worked, than about that of his more conventional predecessor Maximilian II, who was the original owner of most of the pictures. There are references to Tycho Brahe and John Dee, to Paracelsus and Giordano Bruno, as if their ideas necessarily had some bearing on Arcimboldo’s paintings just because their ideas had a certain currency at the imperial court. An elaborate superstructure of artistic theory is even imposed on Arcimboldo’s only recorded comments on art, which come in a letter in which he suggested that the processes of silk manufacture would be a suitable subject for painted grotesque decoration. Yet his remarks seem entirely conventional, and his iconographic proposals are remarkable only for their banality. It is as if the artist is being praised for the books he might have read and the writers he might have met, rather than for his witty and original, if sometimes slightly repellent, pictorial inventions.

The notion that these pictures are derived from the culture of the Hapsburg court is highly dubious and has never been substantiated. But they can readily be related to artistic ideas current in Italy in the middle decades of the sixteenth century, and the only contemporary discussion about them is by Italians. Moreover, it is entirely possible that Arcimboldo had begun to produce composite heads before he left Milan in 1562. Although the Elements have always been dated to 1566, because this date is written on the back of Fire, in origin they probably predate the Seasons of 1563. It is well known that the Elements are mentioned in a sonnet by Lomazzo, which was published, with several hundred others, in 1587; but it has not been noticed by students of Arcimboldo that this sonnet, like most of the others in the same volume, was composed many years earlier. It is cited in Lomazzo’s first literary work, an unfinished series of dialogues called Gli Sogni e Ragionamenti (The Dreams and Conversations), and the reference appears in a section of the manuscript written no later than January 1563. Lomazzo therefore already knew about the Elements by this date. It is just possible that these pictures were begun immediately after Arcimboldo left Milan in 1562, but this is unlikely, especially since the artist does not seem to have kept Lomazzo informed about his activity while he was at the Hapsburg court. This is implied by the fact that Lomazzo’s treatise on art, published in 1584, includes references only to the Elements and a painting called Agriculture (which was probably not by Arcimboldo at all), whereas his Ideal del tempio della pittura, which appeared in 1591, after he had met Arcimboldo again, contains a full account of the artist’s work from 1563 onward.

A further reason for supposing that Arcimboldo had already begun to paint composite heads before he left Milan is the fact that one of his first imitators, Carlo Urbino da Crema, also worked in the same city. Lomazzo tells us in his treatise that Urbino did a painting, Cookery, composed of kitchen utensils, which probably corresponds to a print published in Venice in 1569. A very similar engraving, Agriculture, dated 1567, may also be his invention. And it is worth mentioning that the interest in grotesques to which such compositions most obviously relate is reflected even in the poems of Lomazzo. He himself called these grotteschi. Disregarding the strictures of Horace, he composed them like painted grotesques, playfully combining references to events and people from different epochs and cultures in an ingenious and unexpected way that defies any straightforward interpretation.

In the history of European art Arcimboldo occupies a small but not undistinguished place. In an age of great painters he was hardly more than a competent journeyman. His drawings are dull, his inventions for tournaments and pageants not very different from those of other impresarios of court festivals throughout Europe. But his composite heads initiated a new genre, which, although not very significant in itself, continues to exert a bizarre fascination. Arcimboldo is the outstanding exponent of the type of visual paradox exemplified by Rabbit or duck?, that ambiguous image familiar to all readers of E. H. Gombrich’s Art and Illusion. It was the one thing that he did well, and it brought him fame and success. The exhibition at Palazzo Grassi, and even more the catalog, concentrated on the milieu in which these paintings were most highly prized. The catalog is much less informative about the setting in which they originated, Milan in the years after 1550. But in one sense at least the recent presentation of Arcimboldo’s work is not inappropriate. Just as his painted animals and vegetables can be seen as something altogether more significant, as gods and personifications, so now these amusing trifles have been transformed, by sleight of hand, mystification, and pseudomeaningful juxtaposition, into a major cultural phenomenon.

This Issue

September 24, 1987