Picture Titles: How and Why Western Paintings Acquired Their Names

PICTURE TITLES

PICTURE TITLES

How and Why Western Paintings Acquired Their Names

Ruth Bernard Yeazell

PRINCETON UNIVERSITY PRESS

Princeton & Oxford

Copyright © 2015 by Princeton University Press

Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540

In the United Kingdom: Princeton University Press, 6 Oxford Street, Woodstock, Oxfordshire OX20 1TW

press.princeton.edu

Jacket art: corner of Edgar Degas’s After the Bath, ca. 1895.

Digital image courtesy of the Getty’s Open Content Program.

The Title and Landscape with the Fall of Icarus, by William Carlos Williams, from The Collected Poems: Volume II, 1939–1962, copyright © 1962 by William Carlos Williams. Reprinted by permission of New Directions Publishing Corp. and Carcanet Press Limited.

All Rights Reserved

LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA

YEAZELL, RUTH BERNARD.

Picture titles : how and why western paintings acquired their names / Ruth Bernard Yeazell.

pages cm

Includes bibliographical references and index.

ISBN 978–0-691–16527–1 (hardcover : alk. paper) 1. Titles of works of art. 2. Painting. I. Title.

N7560.Y43 2015

702.8—dc23

2014035489

British Library Cataloging-in-Publication Data is available

This book has been composed in Adobe Caslon Pro

Printed on acid-free paper. ∞

Printed in Canada

1 3 5 7 9 10 8 6 4 2

The Title

—as in Gauguin’s The Loss of Virginity—

how inessential it is to the composition:

the nude body, unattended save by a watchful

hound, forepaw against the naked breast,

there she lies on her back in an open field,

limbs quietly assembled—yet how by its

very unrelatedness it enhances the impact

and emotional dignity of the whole …

—William Carlos Williams, Pictures from Brueghel and Other Poems (1962)

In art, the problems of language are never really settled: language always circles back on itself. Hence it is never naïve (despite the intimidations of culture, and above all, of specialist culture) to ask, in front of a canvas, what it represents. Meaning sticks to man: even when he wants to create non-meaning or extra-meaning, he ends by producing the very meaning of non-meaning or of extra-meaning….

Twombly faces the problem squarely, if only in this: that most of his canvases are titled.

—Roland Barthes, The Wisdom of Art (1979), trans. Richard Howard

CONTENTS

ILLUSTRATIONS

PICTURE TITLES

PROLOGUE (THIS IS NOT A TITLE)

Ceci n’est pas une pipe: This is not a pipe. Perhaps few words more readily summon up a painted image than does René Magritte’s paradoxical commentary on the nature of representation. Both the pipe and its legend have become familiar icons of modern culture, appearing on everything from T-shirts and perfume bottles to the cover of an influential book by Michel Foucault. Yet the picture that so many people think they know does not, strictly speaking, exist. Though Foucault evidently titled his own work after the object that inspired it, there is in fact no painting called Ceci n’est pas une pipe. The photograph that Foucault reproduces with that caption does closely resemble Magritte’s first painting of the subject, but according to the catalogue raisonné of the artist’s work, that painting was completed in 1929 (not 1926, as Foucault has it) and was named by its creator La trahison des images: The Treachery of Images (fig. P-1).¹ The five volumes of the catalogue record many versions of this treacherous image—in this, as in so much else, Magritte was happy to repeat himself—but none of them bears the title with which it is routinely associated.²

I say that none of them bears the title in question because that is a common idiom for the relation of a painting to the words that identify it. But the question of where, if anywhere, we should expect to find such a title is by no means a settled one. Though modern publishers typically place a title below the picture it identifies (as Foucault’s do with Magritte’s painting) and museums often place a label somewhere on the wall adjacent to the painting itself, both of these practices are subject to considerable variation; and the latter in particular, as we shall see, is of quite recent date. (It was still rare, for example, when Magritte first produced his famous image.) To adopt a useful term from Anne Ferry’s study of poetic titles in English, paintings lack a title space: a fixed location where the viewer of a picture can expect to find the words that name it.³ More immediately to the point, the titles of most easel paintings—the original version of La trahison des images included—are not physically part of the work they name. Though Ferry makes clear that the convention of placing a title above a poem has a history of its own, that convention was pretty much stabilized by the culture of print in the late seventeenth century; and the fact that both poem and title inhabit the same medium obviously strengthens the connection. Even when an artist has painted a title directly on the canvas, reading the words remains different from looking at the image.

P-1. René Magritte, La trahison des images (1929). © 2014 C. Herscovici/Artists Rights Society (ARS), New York. Oil on canvas. Los Angeles County Museum of Art. Digital Image © Museum Associates/LACMA. Licensed by Art Resource, NY.

That Magritte did inscribe the words Ceci n’est pas une pipe in large characters across his painting no doubt explains why so many people misidentify it. Those who make the error might find further justification in the fact that the first reference to the authorized title did not appear until a letter from the artist of 1935—six years after the work itself was completed. Though in most cases Magritte followed the practice of writing a title on the back of a canvas, the catalogue raisonné records no such inscription for the original version of La trahison des images. As early as January of 1929 the artist had published a small drawing of the idea in a French journal, but this consisted simply of a pipe (facing the other way) with the words Ceci n’est pas une pipe written below; and when Salvador Dalí described the finished work for a Barcelona newspaper later that spring, he also invoked only the pipe and its apparently self-negating legend. Not until Magritte dispatched a second version of the image for a New York show—this one with the legend painted in English and the title written in French on the back—did the artist apparently refer to the earlier painting, too, as La trahison des images.⁴ Whether he had always thought of it by that name is not clear. By the time that the second version appeared in New York, it would hardly be surprising if some people assumed there was already a picture called Ceci n’est pas une pipe.⁵

Does it really matter what we call Magritte’s famous painting? This book will contend that titles always make a difference, but some titles clearly make more of a difference than others. Indeed, the very fact that Magritte includes some words in his painting—not to mention the force of the proposition they articulate—makes La trahison des images a special case for the problem examined here, if not perhaps for the history of image making more generally. For it is above all when individual pictures appear to float free of words that the question of what to call them becomes most salient. W.J.T. Mitchell has argued persuasively that no image can ever be wholly independent of verbal context; and this book takes for granted that he is right.⁶ But an image that comes physically embedded in such a context—an illumination in a medieval manuscript, say, or an illustration in a nineteenth-century novel—may not even require a separate name; and any name that does identify it will have less effect than it would if that same image were extracted from its context and circulated independently. The case of pictures that include words is less obvious; but all things being equal, they too will depend less on their titles than will those from which all verbal characters have been excluded. (I except for the moment the artist’s signature.) Sometimes, of course, the words within the picture are its title—whether because we know that the artist intended them as such, or because no tradition of a separate title exists.⁷ This book will conclude with a series of works by Magritte’s American successor, Jasper Johns, that deliberately exploit such a possibility by making their own titles integral to the painting. Had Magritte’s letter of 1935 been lost, the canvas that currently hangs in the Los Angeles County Museum might well be labeled Ceci n’est pas une pipe.

As an artist who combined words and images, Magritte long appeared an exception to the standard account by which modern art had progressively emancipated itself from language. Both practice and theory have increasingly challenged that historical trajectory, as the development of so-called conceptual art and the experiments of groups like Art & Language have helped to undermine a narrative whose end point was the will to silence of mid-twentieth-century abstraction.⁸ Beginning with Tom Wolfe’s broad satire on modern art in The Painted Word (1975), commentators have also noted that even that purported silence was inevitably belied by the theorizing that accompanied it, as the words ostensibly banished from the image itself returned with a vengeance in the writing it generated.⁹ However abstract the work of art, Philip Fisher has observed, it is seldom free of the splash of modern explanation which is, in effect, a very long title that the work carries with it, a title known to some and not to others.¹⁰ This aptly summarizes one important way in which images are still dependent on a verbal culture.

Yet the difference between such coterie knowledge and the picture’s actual title is instructive. Unlike the extended account that the work figuratively carries only to those already in the know, the title’s shorthand typically does travel with the picture—at least under modern conditions of reproduction and display—and it is ordinarily available to any viewer. Historically speaking, in fact, the need for titles first became pressing when large numbers of paintings began to circulate, and when local knowledge of one sort or another could no longer be taken for granted among the increasingly heterogeneous public who came to view them. For such a public—which at one time or another includes all of us—the title of a painting often provides the first and even the only language by which the image will be construed. Everyone who has visited a museum or a gallery will know the curious sensation of moving on from one painting to the next and almost before taking in the new image at all, of finding the eyes suddenly plunging down to the tiny rectangle of lettering below or beside it, Norman Bryson has written.¹¹ In this sense, as I shall argue, the history of the title belongs to the democratizing of art—even, or especially, when painters resist the legibility it threatens.

But just what counts as a title? In the last decades of the twentieth century, a few philosophers and other writers tried to wrestle with the problem, mostly by generalizing across the arts.¹² Though their answers vary, the usual impulse is to bring some order to the investigation by ruling certain possibilities out of court. Thus an editor of the Journal of Aesthetics and Art Criticism contends that the unique purpose of titling is hermeneutical and chooses to view as legitimate only those titles that refer to exhibited features of the work—a definition that rules out those that just name the place where a work originated, like Mozart’s Prague symphony, or those that merely identify it in sequence, like Jackson Pollock’s Number 1A, 1948. According to this formula, a title relates to the work itself—rather than to the circumstances of its creation … or to the circumstances of its display or performance. Titling, the author sensibly observes, permits discourse about artworks, but since he limits such discourse to interpretation only, neither words identifying a medium (watercolor) nor those merely naming a genre (sonata, ode) make the cut.¹³ Nothing is said of landscape or still life, but presumably these, too, should not be regarded as titles.

Others want to draw the line not at reference but at authorship. For those who argue in this vein, only artists are entitled to title their work—the etymological connection is part of the argument—and any title thus bestowed is both integral to the work in question and a reliable guide to its meaning. There are titles and titles, in other words, but we should confine our talk to what one philosopher calls "true titles—those given by the artist at roughly the time of creation or constitution of the work" (emphases in the original).¹⁴ Such a stipulation certainly clears a lot of ground, and it has the obvious appeal of keeping the focus on the documented intentions of a work’s creator. It might indeed prove useful, if by no means unproblematic, for distinguishing among literary titles since the age of print. But like other rules that betray an unconscious bias toward literature—the claim that titles appear at the commencement of an artwork, for example¹⁵—this identification of artists and authors is misleading. In the case of painting, I would contend, the purity of such theorizing is achieved only by sweeping aside the complications of history—both the often tangled history by which most paintings that hang in our museums actually acquired their titles and the history that still attends the baptizing of many a modern canvas. At least since the eighteenth century, when the circulation of individual pictures to an increasingly wide and diverse public first made the need for titles salient, there have indeed been cases that would pass the test; and this book will examine some of them. But the true title of a painting, as we shall see, remains as often as not a phantom.

What should we call La trahison des images, for that matter, if the artist’s first use of that title was not recorded until six years after he began to paint such images? Though the problem posed by this particular painting may seem anomalous—few pictures, after all, are inscribed with a memorable competing tag like Ceci n’est pas une pipe—neither the difficulty of determining exactly when Magritte decided to name his picture nor the persistence with which others choose to rename it is exceptional in the history of Western painting. Indeed, for much of that history paintings didn’t have titles in a modern sense; and when the conditions of display and marketing began to require them, they were often composed by persons other than the artist, after the fact. Even when painters themselves more routinely took over the practice, titling remained as much a function of responding to pictures as of creating them, and everyone from dealers and publishers to critics and ordinary viewers got into the act. Sometimes artists openly invited their friends to look at a painting and help decide what to call it—a game that Magritte and his fellow surrealists liked to play, though there is no evidence that they did so for La trahison des images.¹⁶ Nor have such games ended with the digital age, as anyone who has been tempted by Google’s invitation to help label the images that circulate on the Web can testify. While there are clearly exceptions to the rule, the baptism of a painting is apt to be a messy affair: post hoc (sometimes very post hoc), informally negotiated with the artist’s public, and even repeatedly renegotiated as the image itself travels from one context to another. Any theory of picture titles, in other words, will always find itself entangled in a history of reception.

Like paintings, literary works didn’t always have titles, and the titles they did possess didn’t always originate with their creators. Before the seventeenth century, as Anne Ferry has shown, the short poem in English typically acquired its title not from the poet but from the person who presented it to the reader. Both in the manuscript tradition and in the early culture of print, the act of titling was customarily reserved for those I would term middlemen: in Ferry’s taxonomy, such a person might be copyist, editor or commentator, translator, printer, publisher, or bookseller.¹⁷ Ferry credits Ben Jonson with being the first poet in English to take the titling of his poems seriously, though authors of longer works had surely been composing titles well before the age of print. (It was Dante who called his long poem the Commedia [c. 1314–21], even if the adjective Divina is a later addition that first appeared on a printer’s title page in the mid-sixteenth century.) Throughout the first half of the seventeenth century, Ferry suggests, poets could exploit the remaining uncertainty as to the status of their titles by creating the fiction that someone else was responsible for presenting their verse—a fiction that continued to influence the grammatical form of many a title long after the question of authorship was more or less settled.¹⁸

Ferry’s observations as to how poets’ titles often follow the conventions originally adopted by middlemen (referring to the speaker of the poem in the third person, for example) have significant resonance for the practice of picture titling, as we shall see. But for now it is the difference between picture titles and the literary kind that I want to emphasize. The most obvious, of course, is the difference in medium. While Ferry doesn’t say so explicitly, our working assumption that the poem and its title have the same author is surely facilitated by the facts that both are made of words and that they customarily appear as contiguous marks on the same sheet of paper. In the case of longer works, the title may seem somewhat more divided from the text itself by the presence of the title page, but the very standardization of print that produced such a page in the first place partly overrides that distinction:¹⁹ by concealing any signs that title and text might be the products of different hands, print encourages us to approach them as the continuous creation of a single mind. Many books probably have owed their titles more to their publishers than to their authors, but the appearance of the final product has made this division of labor relatively easy to ignore.²⁰ The development of the printing press also meant that both work and title could be multiplied indefinitely, and that they could circulate together with relative ease. With the partial exception of the reproductive print—an object that is not of course the painting itself but its translation into another medium—and those special cases where the title of the picture appears to be inscribed on the canvas by the hand that created it, none of this holds true for painting. And even when a picture and its title are physically united in the same object, the difference in medium still demands different kinds of reading.

P-2. Pietro Martini, Exposition au Salon du Louvre en 1787. Engraving. © Trustees of the British Museum, London.

For much of its history, in fact, the individual oil or watercolor was unlikely to be displayed in close proximity to its title. Contemporary museums have accustomed us to look for the label on the wall immediately adjacent to the painting, but such exhibition styles are of quite recent vintage. Throughout the eighteenth and nineteenth centuries, the pictures on display in public galleries and other venues were typically hung frame-to-frame, with no room for commentary of any sort on the walls themselves: to identify the image verbally, one consulted a catalogue. (Figure P-2 depicts the Paris Salon in 1787, and figure P-3 shows the exhibit of the same year at the Royal Academy in London.)²¹ Although aesthetic pioneers like James McNeill Whistler and the owners of the Grosvenor Gallery in London began to pare down this crowded hanging style in the 1870s, the practice lasted well into the first decades of the twentieth century: even the famous show of avant-garde paintings at the New York Armory in 1913 was hung accordingly. Not until the Museum of Modern Art in the same city helped to establish more spacious hangings as the norm did the custom of providing wall labels become standard, though some exhibits, including the 1929 opening at MoMA itself, continued to reserve titles and other verbal information for the catalogue.²²

P-3. Pietro Martini after Johann Heinrich Ramberg, The Exhibition of the Royal Academy, 1787. Engraving and etching. © Trustees of the British Museum, London.

Unlike literary titles, in other words, picture titles long appeared at a certain remove from the works they named—a remove at once physical, as in the gap between painting and catalogue, and often temporal too, as in the countless works that did not acquire the titles by which they were sold or displayed until years and even centuries after they were painted. By the end of the eighteenth century, artists were expected to accompany their submissions to the Paris Salon or the Royal Academy with some identifying language for the catalogue, but the young woman who consults her copy in the doorway of figure P-3 might well be forgiven for thinking of such language as more like a brief interpretation of the painting than a part of the work itself.

At the same time, such an interpretation was likely to shape her experience of the image in a way that few literary titles could match. We have learned to be skeptical of Gottfried Lessing’s famous distinction between the temporal and the spatial arts, but precisely because the temporal sequence by which we view a painting is not conventionally structured—because the words that name it do not come at its commencement—the title of a painting retains a greater power to shape our experience of the whole.²³ The words of a literary title are succeeded by other words that may diminish or efface them—many words indeed in the case of longer forms like novels or plays—but the words of a painting’s title typically stand alone, with little other than the artist’s name to qualify or counter their effect. Of course, this difference between picture titles and the literary kind is not absolute: in the eighteenth and nineteenth centuries, catalogues like the one being read by the young woman in Martini’s print sometimes followed up their titles with extracts from poetry or other texts, while in our own day museums provide expository labels for selected works on display. And technologies for replicating images mean that a version of the painting, if not the painting itself, can theoretically be accompanied by a verbal explanation of any length, whether in a book containing a reproductive print or photograph, or digitally on a screen. But to the degree that the title serves as the invariant shorthand for the image wherever it travels—a shorthand that persists through accounts that otherwise may vary radically—its words will always retain their privileged status. Until visual recognition software becomes as accurate as the verbal kind, people who search for a particular painting on the Web will continue to do so by its title. Art historians who thoroughly disagree about the meaning or significance of a particular painting will continue to call it by the same name. And even when they seek to argue that a painting should in fact be designated differently, its traditional title, as we shall see, may prove stubbornly resistant to change.

In 2009 the National Gallery in London organized an exhibit designed to highlight Pablo Picasso’s continuing dialogue with the masters who had preceded him. Commenting on the exhibit for the London Review of Books, the journal’s regular art critic, the late Peter Campbell, singled out for particular notice a cubist work of 1910 (fig. P-4). Without the title, Campbell wrote, "you would be hard pressed to relate any detail of Still Life with Glass and Lemon to an object. Only after you have decided that there is a central stand or table and objects on it can the game of interpretation begin. Old art is a substrate, but in this case the viewer borrows an idea to look with rather than the painter an idea to build on."²⁴ Campbell was surely right about the force of this title, though his distinction between the viewer’s idea and the painter’s tellingly equivocated about its status, as if he couldn’t quite believe that the way in which we are invited to classify the painting corresponds to the original intention of its maker. So great is the apparent distance between this severe instance of analytic cubism and the pictorial tradition to which the title alludes that the critic half suspects the latter of being an afterthought.

P-4. Pablo Picasso, Still Life with Glass and Lemon (1910). © 2014 C. Herscovici /Artists Rights Society (ARS), New York. Oil on canvas. Cincinnati Art Museum, Bequest of Mary E. Johnston/Bridgeman Images.

Presumably he would not likewise pause to remark the title currently borne by Willem Kalf’s Still Life with Lemon, Orange, and Glass of Wine (fig. P-5), for instance, or the numerous variations on the formula by which so many paintings since the seventeenth century have been identified. Yet these too are ideas to look with, and despite their seeming self-evidence, they also have their place in a theory of picture titles. Rather than assume once again that the story of modern painting begins when the French impressionists break with the Académie in the middle of the nineteenth century—the assumption that governs the only substantial work on picture titles we have to date²⁵—this book contends that we need to look further back, to the anonymous classifications and other names under which large numbers of easel paintings began to circulate in the late seventeenth and eighteenth centuries. It is not enough to say, with Arthur Danto, that the history of titles before impressionism is simply the history of motifs. Danto suggested that titles become interesting for art history only when they begin to call attention to certain features of painting as painting²⁶—when Whistler, for example, chooses to call his portrait of his mother Arrangement in Grey and Black (see fig. 14–4). But a generic title like still life or landscape is also a way of attending to painting as painting; and so too, for that matter, is the portrait that subtitles Whistler’s painting of his mother. The degree to which we have come to take such categories for granted should not obscure their capacity to shape the viewer’s experience of an image.

P-5. Willem Kalf, Still Life with Lemon, Orange, and Glass of Wine (1663–64). Oil on canvas. Staatliche Kunsthalle, Karlsruhe.

Nor is the naming of motifs ever as simple—or as visually innocent—as the dismissive generalization implies. In a related argument, Leo Hoek has characterized many such titles as merely denotative and hence transparent, but the very act of singling out some features of the image has an effect, whether or not the viewer consciously pauses to register it.²⁷ Portrait of the Painter’s Mother is not the same as Portrait of an Old Woman; and the effect would change yet again if we just subtitled the work An Old Woman—or, in a subtly different move, The Old Woman. In that case, of course, we would have turned Whistler’s portrait into a genre painting (though the definite article might render that classification slightly ambiguous)—a gesture we may well wonder that the artist himself didn’t make, since he later insisted that his personal feelings for his