Meyer Schapiro, who could plausibly claim to be the most interesting art historian today, has not yet written a book. This is part of the legend that surrounds him, as well as one of the reasons why he is still a controversial figure. It is not, however, strictly true. In 1964 he published a short monograph on a Romanesque manuscript, The Parma Idelfonso,1 and earlier, in 1950 and 1952, books on Van Gogh2 and Cézanne;3 but excellent as they are and influential as they have been, these last are essentially long essays with notes to the plates. The new little volume under review will not change the situation: for many people, especially envious colleagues, Schapiro will still be the famous art historian who never wrote a book.
Not that he is unproductive. A man of bewildering intelligence, with a formidable memory, the widest culture, and an inexhaustible curiosity, he has since the mid-1920s regularly written articles that range from the most technical subjects to the most general, from Greek art to current movies. His two principal fields of scholarship are astonishingly far apart: nineteenth-century French painting, and the early Middle Ages from Irish and Carolingian manuscripts to Romanesque sculpture; he has also contributed much to the criticism of modern art. His mind is basically a critical one; characteristically, many of his contributions are book reviews, sometimes harsh and devastating ones, but always illuminating the subject.
The underlying principle of his criticism, in this new book as in his other work, is that a single observation can always be interpreted in different ways. Only a variety of observations and arguments can prepare the way for a relatively definite conclusion about a work of art. For Schapiro all possible clues have to be followed up; he draws on a range of knowledge well beyond the usual equipment of art historians and uses it with all the sophistication of traditional philology. “I see that you know very little about the history of dirigibles,” he is supposed to have said, sadly but kindly, to a student who had guessed wrong the date of a painting where such a machine appeared.
Are these philological investigations only a game? Or do they bear upon our understanding of art? For some critics art should speak for itself without learned interference; like Malraux, they bring all works into a universal present. Others, more historically minded, claim that we can understand the art of the past only by a scholarly reconstruction of the original circumstances of its creation and the contemporary attitudes to it. They want to carry us back into the past. The first position may appear as a brutal manipulation of the work, an appropriation of it. It severs the work from the intentions of the artist and from the conditions of its appearance, makes it available only for consumption. The second, however, is reactionary escapism. One recalls, for instance, the attitude of Dietrich von Bothmer at a particularly unfortunate moment during the controversy over the Metropolitan Museum’s new vase: Why can’t we transport ourselves back to fifth-century Athens? (In other words, why can’t we forget about the present mess?)
Both views try to abolish historic distance: one attempts to erase the past; the other wishes to forget the present. Schapiro is acutely aware of this problem of historic distance: his views on it are complex, not to say reticent. He insists on the value of the innocent eye, the direct encounter with the work of art, and does not believe that an uninformed experience is illegitimate or inferior, or even that knowledge always changes or improves our appreciation. But he also knows that information and study can, at times, contribute to aesthetic understanding, and even improve our actual perception. In Schapiro’s work we also sense the pure enjoyment of scholarship and argument, the pleasure in Talmudic dispute, of a man assertively attached to his Jewish culture and tradition.
It is hardly new to say that art has several kinds of meaning, but Schapiro is the art historian who has taken this polysemic nature of art most seriously, approaching it from many different and unexpected angles. Not only a master of iconography and formal criticism, the traditional tools of art history, he is extraordinarily well versed in a variety of other disciplines. In the prewar years, when some of his most interesting work appeared in The Marxist Quarterly and Partisan Review, his concerns were heavily social and political; in the 1950s he concentrated on applying modern psychology to the artist and his work; more recently he has been particularly concerned with the semiotics of art: its functioning as a language, or system of signs. But none of these interests has excluded the others. If one approach may be more useful in revealing a particular aspect of a work of art, or a particular kind or period of art, it is never exhaustive.
We can perhaps understand why Schapiro has not written a major book: he is reluctant to neglect or renounce anything related to our understanding and enjoyment of art. A book, conceived as a grand encompassing whole, is impossible to write without the kind of sacrifice Schapiro is not prepared to make.
Short as it is, Words and Pictures: On the literal and the symbolic in the illustration of a text has a large subject and a wide range of examples, and it is more complex than most works on art history. Its main theme is the relation of picture to text in scriptural illustrations during the Middle Ages. Schapiro is concerned above all with the different possibilities that exist for the illustrator as well as with the constraints that are imposed on him. He wants to do justice both to the power of the image to reach the innocent eye and to the power of scholarship as an instrument that can elicit meaning.
The artist who illustrates a text cannot render all the ideas expressed in it and is often forced to add circumstances that do not appear in the text. For example in the story of Cain and Abel, the artist must show the murder weapon, which is not mentioned in the Bible. He either can give a strict pictorial equivalent of the text or render it with a symbolic interpretation. The story of Abraham sacrificing Isaac can be represented simply for its narrative value, or the artist can bring into the picture references to the Christian view of the story as prefiguring the sacrifice of Christ. In this case, the pattern of a cross in the wood carried by Isaac can make the symbolic interpretation explicit.
Schapiro deals at length with one example: the illustration of a text from Exodus, the story of the battle against Amalek, especially the verses,
- And when Moses held up his hand, Israel prevailed; and when he let down his hand, Amalek prevailed.
- But Moses’ hands were heavy. So they took a stone, and put it under him, and he sat on it. And Aaron and Hur stayed up his hands on both sides. And his hands were steady until the going down of the sun.
Schapiro first shows us that under the Carolingian empire a new image was used for this episode: a frontal presentation of Moses with his arms propped up by Aaron and Hur, a design suggesting the figure of the Cross. He then discusses the later development and eventual abandonment of this image. His concern is with the ways in which historic conditions, the choice of a particular symbolic interpretation, and the artistic means used to convey it are all related.
The only known early Christian representation of this subject (Rome, Sta Maria Maggiore) dates from the fifth century. Moses stands as a praying figure with both arms raised but unsupported, making the sign of the Cross. It is not consistent with the text, but it satisfied the demands both of Christian symbolism and the classical stylistic norm of a free-standing heroic figure that still prevailed at the time.
The new medieval image in which Moses’ arms are supported by Aaron and Hur appears around 880. It cannot be explained simply by a fresh reading of the text. Innovation was not a simple matter in the Middle Ages; for a new image to establish itself required not only a change in aesthetics and outlook, but also appropriate historical circumstances. Here Schapiro discovers a link with the new structure of power in the Carolingian empire; he shows us images in which the emperor’s arms are similarly supported and where the artist was probably making a reference to Moses. (On the other hand, it is curious to see that in a Jewish manuscript the repellent symbolism of the Cross is rejected.)
Both the frontal view of Moses and the symbolism of the Cross it carried with it were later abandoned. Schapiro relates this change to a general shift in art in the twelfth and thirteenth centuries, when static representations were replaced by more animated ones that favored the profile or semiprofile. This change, he writes, is characterized by a “heightened interest in action, whether in religious or secular scenes, as an objective engagement.” Eventually, the battle against Amalek was sometimes represented merely for its interest as an epic, with Moses in the background or even left out altogether. And Schapiro finds comparable changes in biblical exegesis at the time.
The significance of the frontal view of Moses suggests the general problem of “frontal and profile as symbolic forms,” which Schapiro deals with in his final chapter. Schapiro treats the frontal and profile positions as an opposed pair that can assume a variety of meanings. In the Middle Ages the fullface expresses the hieratic or the stately (used for God or Christ) while the profile is often reserved for the evil and the outcast (Judas in the Last Supper). But this is not always true during the medieval period, and at other times the meaning can be completely different. Schapiro mentions, for instance, a Greek vase where a mother and her child are shown in profile while the servant who carries the child is in fullface. Profile and fullface do not have a stable established meaning: “The contrast as such is more essential than a fixed value of each term in the pair; what counts is the distinction of rank by a different relation to the viewer.”
Such a statement suggests that the meaning of art is conventional or arbitrary—i.e., that there is nothing “inherent” or “natural” in the meaning we give such positions—and it points to a consideration of art as a language, justifying the place of Schapiro’s essay in a series called Approaches to Semiotics. Schapiro had already explored the study of art as a system of signs in a fundamental essay, “On Some Problems in the Semiotics of Art: Field and Vehicle in Image Signs.”4 There he called attention to certain basic elements of art which often go unnoticed, like the smooth surface of the picture, the limits of the picture field, the frame, the opposition of right and left parts of the field and of the upper and lower ones. He showed how such elements first appear as innovations, become adopted as part of artistic practice, and can carry a variety of meanings (e.g., right is good, left is bad).
But even in this essay Schapiro was reluctant to isolate the systematic aspects of artistic creation. He always remained aware of the way art is rooted in life and of the dangers of a formalist history that considers art only in itself, depending on its own momentum, its own rules and development. Consequently Schapiro is concerned both with the relation of art to other aspects of human activity and culture (ways of life, ritual, politics, or even psychic life) and with the possibility that some natural relation exists between the symbols of art and their meanings.
Is there, for instance, some quality inherent in the profile and the fullface that predisposes them to symbolize one thing rather than another? The fact that some elements can carry different and even opposed meanings at different times in history is a traditional argument for the conventional or arbitrary nature of artistic meaning. As Schapiro writes, “The plurality of meaning in each of these two appearances of the head would seem to exclude a consistent explanation based on inherent qualities of the profile and the frontal or fullface view.” He seems here to tend toward a view favoring purely conventional meaning.
When he turns to the symbolism of color, however, Schapiro takes a somewhat different view. “The familiar argument,” he writes, “that color symbolism is entirely conventional, ignores that color is not a simple elementary feature but a complex of qualities of which certain ones become more or less pronounced in a particular setting and according to a perceiver’s experience and attitude…. By its purity and brightness white can symbolize the spiritual and innocent; by its lack of color, the inert and cold.” Although this defeats the historical argument in favor of “pure” artistic convention, it seems no more than tentative in finding that it has some basis in nature.
If Schapiro’s argument here seems inconclusive, it may be that the question should be put a little differently: Is it possible to study art as if it were a system of conventions? Certainly language can be studied and described as such a system, and this has helped us to understand how it functions, although like color and the forms of art, language, too, has aspects that are not exhausted by such a study. In fact all the efforts since the nineteenth century to make the study of art more “scientific” have been accompanied by a comparison between art and language; but all such comparisons have almost always remained no more than loose metaphors. Schapiro himself makes frequent comparisons with language in his last chapter, and even there it is not always absolutely clear how these comparisons are to be taken, although it is clear that he is aiming toward a semiotics of art.
Semiotics as the general science of signs is an idea or an ambition largely inspired by Charles S. Peirce and Ferdinand de Saussure which has attracted more attention in recent years. It has always seemed obvious that the study of art would legitimately belong to it, but at the same time very little progress has been made toward any convincing system of artistic signs. One of the main difficulties is that since language is the most important and the richest of all systems of signs, linguistics appears as the necessary model of semiotics.
But there is a serious obstacle to any attempt at establishing a semiotics of art on the model of structural linguistics, which takes as a working hypothesis a distinction of the kind that Saussure made between language (la langue) as opposed to speech (la parole). Language here is a system that governs linguistic usage at a given time and can be described independently from particular statements or texts. Although this is an abstraction, it has proved to be a workable one. It assumes, however, that one studies the whole of language as one system so that its elements take their meaning from their place in the system, just as the value of a piece in a game of chess does not exist independently of the entire set of rules.
The symbols of art are of a different kind. Schapiro, for example, compares the profile position to the deferential use of the third person pronoun (“If Madam wants me, she will find me in the garden”). We can understand this only as a suggestive way of describing the meaning of certain profiles we see, but not the way the meaning is actually conveyed to the viewer.5 The profile, in so far as it symbolizes a particular relation between the viewer and what is represented, does so metaphorically through a resemblance or analogy between what we see in the picture and our experience of the world. When we see actual people in profile no psychological exchange can take place through a glance, and so we experience a sense of distance from them.
Any discussion of the conventions in works of art points up in a similar way their sustained metaphorical quality, their constant analogy with experience as well as the analogies they evoke within the work itself. We say that the Virgin’s mantle is blue like the heavens. When the profile is an attribute of evil we see it as a ragged line. When it ennobles Renaissance portraits, like those of the Duke and Duchess of Urbino by Piero della Francesca, it does so of course partly by recalling ancient coins and medals. But there is also an inherent quality of the profile that comes into play. When we see a person in strict profile, changes of expression are more difficult to observe than in the three-quarters or frontal view, and they alter the features much less. It is this immutability, this permanence of the profile that is brought out in the portraits. One constantly discovers or invents a natural basis for conventions by drawing such analogies. This way of bringing out metaphorical qualities in perceiving art, incidentally, does not apply only to representation; abstract ornament and modern non-representational art seem to me to function in the same way.
Since the symbols of art are always capable of suggesting an analogy or a metaphor to support their meaning, their dependence on any system of conventions is loosened, and this seriously limits the possibility of constructing a complete or coherent artistic code for a given time. The coherence of the single work may seem so strong and binding that we tend to assume that the work of art establishes its own code and language. But this is only true up to a point. Linear perspective, for instance, is a convention and a code. This does not, however, exclude the possibility that there is a natural basis to the convention or deny that it may be more “true” than other modes of representation; the latter question must depend on the auxiliary question: true to what? A work of art can only elaborate—and occasionally innovate—on the basis of existing forms and conventions, even if it sometimes gives the impression of reinventing them, as poetry can give the impression of manipulating or re-creating grammar.
A deeper and more inclusive investigation of the conventions of art is badly needed, notwithstanding all the difficulties I have mentioned, and here Schapiro’s book has much to suggest. One of his most pregnant ideas, for example, is that the range of meaning of the opposed pair—fullface/profile—as a symbolic form partially depends on the mode of representation current at different times. In some periods, as we have seen, the profile or the frontal view may be predominant, but at other times, particularly in more naturalistic styles, these positions can be exceptional. In the Middle Ages both were exceptional, while a stereotyped three-quarters view was normal. Clearly the use of these symbolic forms was bound by a system, even if it was a rudimentary one.
A statement of Schapiro’s leads us a bit further in understanding how this system worked: the stereotyped three-quarters view, he writes, “combined aspects of fullface and profile, independent of an explicit connection of the eye or hand with a specific object.” In other words, the turn of a figure’s head is largely independent of the figure’s action. In medieval pictures, an explicit connection with an object could be indicated by a pointing finger but also by the eye, that is, the glance.
This, however, implies another conventional device: the glance is not indicated by the direction of the eyes in a realistic manner, but by an invisible line that joins the object to the axis of the figure’s two eyes. In other words, we are to understand that a figure is looking at an object only if the object is lined up in the picture plane with both eyes of the figure. By placing the pupils at the side of the eyes in the direction of the object, medieval artists gave a more vivid conviction to this arbitrary convention. Even though a figure in a medieval illustration may appear to be looking into our space, this convention implies that it can look only within the picture plane. The relation between the position of the head and the direction of the eyes makes the conventional system more complex.
It is disappointing that Schapiro, who is obviously aware of this device for indicating the glance of figures in the standard three-quarters view, does not mention it. For both the fullface and profile positions of the head lack this ability to indicate direction. The fullface always stares at the viewer wherever he may stand, so that it gives the impression of looking everywhere at the same time. Schapiro remarks on this quality; he quotes medieval texts that show an awareness of it and relate it to the omnipresence of God. The profile, on the other hand, showing only one eye, cannot obey the same convention as the three-quarters view, but like the device of the two-eye axis, it encloses the direction of the eyes within the picture plane and picture world; the eyes cannot look out of the picture so as to involve the viewer.
It is only at a later stage, in the late thirteenth century, that the direction of the glance becomes more flexible and is rendered as independent from the position of the head in the picture plane but consistent with the direction of the head in a projected three-dimensional picture world in which the viewer can be included. Only then can the figures look in any direction, inside as well as outside the picture.
The emotional power of Giotto’s Betrayal of Christ, as Schapiro points out, depends on a new use of the profile. Here the contrast, traditional for this scene, between the profile of Judas and the fullface of Christ, is replaced by two contrasted profiles that confront each other. “In Giotto’s painting all the figures behind Christ and Judas, except for the profile heads of Peter and Malchus, which reenact with a reduced intensity the opposition of Christ and Judas, are turned towards the center and reinforce the intent glances of the main actors.” All this, however, is only made possible by the consistency of the direction of the heads with a strongly indicated glance. These devices in Giotto are rudimentary and rough but very powerful: they reveal the experimental stage of a wholly new form of representation, and the excitement of a new view of reality. At the same time, we see a new pictorial order where the arrangement of the volumes overrides the linear patterns on the surface of the picture—which were still dominant some ten years earlier in the Betrayal scene of the upper church at Assisi. As Schapiro makes clear, the new style and the new human drama epitomized by Giotto in the opposed profiles of Judas and Christ cannot be separated from each other.
Schapiro’s analysis of the semiotics of the profile/fullface is one of the most suggestive of such studies ever to appear. Presumably it could be carried further but this probably would also mean narrowing the point of view. For Schapiro also makes us aware of the relative independence of artistic symbols from any tight system or code—an independence due, as we have seen, to the way they behave as metaphors. Even a feature as fundamental and widespread as the frontal/profile opposition is utterly entangled both in history and in the context of the single work. This emerges clearly from the central chapters of Words and Pictures, to the richness of which I have hardly done justice. They prepare us for the conclusions of the last chapter on frontal and profile forms, but also act as a powerful qualification to any schematic generalization.
Schapiro’s contribution to the semiotics of art, a subject that is now attracting more and more interest, is especially important not because of his interpretations of individual works—in spite of his brilliant insights—but because he illuminates the conditions from which the multiple meanings of art emerge. He shows that by analyzing the place and function of an element in the different sequences or systems it belongs to (such as the structure of the work, a code of conventions, a historical period, the psychological projections of the artist) its different meanings can be made explicit. The work of art is too often presented as made up of two separate things, a formal order and a subject matter, which, in spite of many pious words on the necessity of unifying them, are kept neatly apart. Schapiro’s great accomplishment has been to break this unfortunate mold, showing us the way out of an impasse in art history.
If, however, one has a serious complaint about Words and Pictures it is its brevity. One would like the aperçus and qualifications to be more fully developed, more explicit. At points Schapiro’s laconic text makes him seem too reticent. Perhaps he has to be.
MEYER SCHAPIRO: A SELECTED BIBLIOGRAPHY
(Omitted here are titles mentioned in the accompanying review.)
I. On General Aspects of Art History
“Style,” Anthropology Today, edited by A. L. Kroeber (Chicago, 1953). Reprinted in Aesthetics Today, edited by M. Philipson (New York, 1961), pp. 81-113.
“Race, Nationality and Art,” Art Front, 2, March, 1936, pp. 10-12.
“On the Relation of Patron and Artist: Comments on a Proposed Model for the Scientist,” The American Journal of Sociology, 70, 3, 1964, pp. 363-369.
II On Medieval Art
“On the Aesthetic Attitude in Romanesque Art,” Art and Thought, edited by K. B. Iyer, issued in honor of Amanda K. Coomaraswamy (London, 1947), pp. 130-150.
“The Sculptures of Souillac,” Medieval Studies in Memory of A. K. Porter, edited by W. Koehler (Cambridge, 1939), Vol. II, pp. 359-387. Also published separately: The Sculptures of Souillac (Cambridge, Mass., 1939).
Review of Kurt Weitzmann, The Fresco Cycle of S. Maria in Castelseprio, Art Bulletin, 34, 2, 1952, pp. 147-163.
“Cain’s Jaw-bone that did the first murder,” Art Bulletin, 24, 1942, pp. 205-212. See also pp. 383-384 for a letter response from A. K. Coomaraswamy.
“The Religious Meaning of the Ruthwell Cross,” Art Bulletin, 26, 1944, pp. 232-245.
“The Bowman and the Bird on the Ruthwell Cross and Other Works: The Interpretation of Secular Themes in Early Medieval Religious Art,” Art Bulletin, 45, 4, 1963, pp. 351-355.
“Munscipula diaboli, The Symbolism of the Merode Altarpiece,” Art Bulletin, 27, 1945, pp. 182-187.
“A Note on the Merode Altarpiece,” Art Bulletin, 41, 1959, pp. 327-328. (This and the preceding essay are reprinted in Renaissance Art, edited by Creighton Gilbert [New York, Harper Torchbook, 1970], pp. 21-42.)
III. On Nineteenth-Century Art
“Courbet and Popular Imagery: An Essay on Realism and Naïveté,” Warburg and Courtauld Institute Journal, 4, 1940/1, pp. 164-191.
“Cézanne’s Apples: An Essay in the Meaning of Still-life,” Art News Annual, 38, 1968.
“The Still-life as a personal object, a note on Heidegger and Van Gogh,” The Reach of Mind: Essays in Memory of Kurt Goldstein (New York, 1967), pp. 203-209.
IV. On Modern Art
“The Nature of Abstract Art,” Marxist Quarterly, 1, 1, January-March, 1937, pp. 2-23. For a muddled response to this article see Delmore Schwartz, “A Note on the Nature of Art,” Marxist Quarterly, 1, 2, pp. 305-310, with a reply by Schapiro following, pp. 310-314.
“Rebellion in Art,” America in Crisis, edited by Daniel Aaron (New York, 1952), pp. 203-242. (On the Armory Show.)
“Gorky: The Creative Influence,” Art News, 56, 5, 1957, pp. 28-33. Reprinted as introduction to L. Goodrich, Arshile Gorky (New York, 1957).
“A Metaphysics for the Movies” (Review of M. T. Adler, Art and Prudence), Marxist Quarterly, 1, 3, October-December, 1937, pp. 406-417.
V. On Critics
“Leonardo and Freud: An Art Historical Study,” Journal of the History of Ideas, 17, 2, 1956. Reprinted in Renaissance Essays, edited by P. O. Kristeller and P. Wiener (New York, 1968), pp. 303-336.
“Fromentin as a Critic,” Partisan Review, 16, 1, 1949, pp. 25-51. Reprinted as an introduction to Fromentin, The Old Masters (New York, Schocken, 1963).
“Mr. Berenson’s Values,” Encounter, 16, 1, 1961, pp. 57-65.
“Looking Forward to Looking Back” (Review of L. Mumford, The Culture of the Cities), Partisan Review, 5, July, 1938, pp. 12-24.
November 14, 1974
The Parma Idelfonso: A Romanesque Illuminated Manuscript and Related Works (New York, 1964). ↩
Vincent van Gogh (Abrams, 1950). ↩
Paul Cézanne (Abrams, 1952). ↩
Semiotica, I (1969), pp. 223-242. ↩
Schapiro’s example, however, is carefully chosen because it is a displaced grammatical form, the origin of which has something in common with the signs of art. However, once it is accepted in linguistic usage, it is a conventional way of indicating a relation between two persons just like other pronouns. ↩