University of Virginia Library

Editing Paintings/Conserving Literature: The Nature of the 'Work'
by
Paul Eggert [*]

Critical editing has often been justified, if sometimes complacently, by its parallel with the restoration of discoloured and damaged paintings. But the complacency is easily pricked by a conundrum editors are fond of quoting: if the Mona Lisa is in the Louvre, where is Hamlet? The question teases the listener into thought about the existential status of the literary as opposed to the plastic arts—of the basically different relationships of the 'work', in each case, to the physical object.[1] The present essay examines some of the practices and assumptions of modern art conservation for the—often refractory—light they throw on traditional editorial beliefs about the boundaries and constitution of the literary work.

I

Among the most important pieces of art restoration in progress at the moment is the work on the ceiling of the Sistine Chapel. Time has wreaked remarkably little havoc since the fifteenth century. Ceiling leaks, candle smoke, and the wine, animal glue, arabic resin and other substances used in the many earlier restorations dating from the sixteenth century are nearly all being removed. The dark and gloomy surface which suffocated the range of Michelangelo's tonality and fed the nineteenth-century myth of his being a black and melancholy artist is being stripped away. Dott. Colalucci, the Senior Restorer, has argued the 'necessity to recover all full chromatic effects intended by Michelangelo . . . without which [the fresco] would appear flat and without modelling'.[2] The aim is a total cleaning. One of the discoveries has been a piece of the original fresco which in 1517 was used to fill a crack which had already developed in the ceiling. The colours of this fragment, which had suffered only a few years of exposure, are extremely close to the colours emerging from the restoration process. Relatively little retouching

has been found necessary, and that has been done in watercolour, thus making it easily reversible by later generations of restorers. One result of this ambitious restoration—spanning nine years—is new knowledge about Michelangelo's working methods. In order to make decisions about methods of restoration, the restorers have had to reconstruct how Michelangelo actually went about his work. The surface of the paintings and the subsurface have had to be investigated thoroughly. Very little evidence of co-workers has been found; the ceiling was not therefore a workshop production. Michelangelo painted up to eighteen square feet per day. Painting a fresco, he had to work while the surface was still damp: it has been calculated that he did the whole ceiling in about 450 stints or giornate, each of up to a day in length.

An interesting parallel with bibliography emerges. The painting, we learn, has had to be analysed not just as image, for its iconic value and historical interrelatedness to the traditions of the period. It has had to be known, in the words of the title of a recent Australian conference of restorers and curators, as an 'articulate surface'. Restorers (or 'conservators' as they prefer to be called) and curators nowadays have available to them a battery of tests: X-radiographs which seek evidence of pentimenti (revisions) or earlier versions beneath the surface, the use of raking, ultra-violet and infra-red light, the microanalysis of cross-sections of paint, and the many other techniques of chemical identification. The building-up of the layers of paint from the 'ground' coating can be examined. These methods could be said to correspond to the collations and the methods of analytical bibliography employed by editors; and descriptive bibliographies could be paralleled to catalogues raisonnés. Indeed, if curators and conservators are archaeologists of the image—if they can make the surface articulate the painting's history—then editors are archaeologists of the printed or written text, of the history of its writing and production. But in both cases it has traditionally been felt that the primary upshot of this considerable effort is to present to the viewing or reading public a restored painting or a reading text of the literary work—the real thing, or as close to it as we can get. Just as, it is believed, accumulations of dirt and varnish can be removed to reveal the painting as it left the artist's hand, so ought editors to aim at recovering the text which the author (however one construes the term) intended. Through the glass darkly, we can—with conservators' and editors' help—espy, if not know in all its details, the thing itself: the work of literary or painterly art.

However, the following consideration of the activity of art-making—and the traces it leaves behind for the conservators to deal with—does not support the traditional assumption. This conclusion might be of only tangential interest for literary editors were it not that the evidence from painting conservation dovetails in some important respects with that provided by editors' examination of the traces of literary production. If the line of observation and argument in this essay is persuasive, then there are implications for editors' understanding of their practice because of the unavoidable questioning

of the traditional term, 'literary work'. Rather than a robust existential entity, it emerges as a creature of our own conventions.[3]

II

As a professional activity, art conservation has emerged only since the Second World War, but in recent years a critical assessment of the theoretical bases of the activity has been in train. This is evident in the papers given by conservators and curators at recent conferences in Ottawa, London and Canberra.[4] Assumptions about the boundaries of the work of art are being contested—as are the roles which the viewing audience and the exhibition context should have in the definition of the work.

Even the case of the Sistine Chapel—which might have been custommade to illustrate the traditional view—has its refractory elements. Minor retouchings from previous restorations are being removed. Dott. Colalucci refers to them as 'unartistic'. But some additions done in 1565 as a result of the Council of Trent's finding the nudity in the painting disgusting are being retained, despite the fact that their removal would be a simple matter. Daniele da Volterra painted some small pieces of clothing over the offending portions. A copy in Naples shows how the painting originally looked in this respect. This, incidentally, is a familiar paradox to conservators: that because of more favourable conditions of hanging or storage, the use by the

copyist of superior materials, or a botched restoration of the original, a copy can be closer in some respects to the earlier state of the painting than is the painting itself.[5] But while the copy may repeat faithfully the original image and still approximate its original colours, its surface and subsurface will almost certainly not correspond. They will articulate a different process of production, a different conjunction of historical factors.

The decision, in the case of the Sistine Chapel, to retain what critical editors might think of as the Council of Trent's bowdlerisation is being done on the grounds of their historical value, despite the fact that the justification of the entire restoration is based on appeals either to Michelangelo's intention or to the recovery of the original state of the frescoes. This does not seem a consistent application of a principle, if the principle is authorial. If the principle is historical, then it is running together two historical moments within a third: the moment of late-twentieth-century restoration. The effects of time on the painting are not being eliminated; they are being packaged in a superb, though somewhat contradictory, act of artistic editing—an act that is happening in the present. Although there has been a long history of failed restorations of the Sistine Chapel and although Dott. Colalucci envisages the need for successors, the rationale of his team's activities is backward-directed. Yet clearly their restoration, which is an intervention of the most radical kind, eliminates the evidence on which generations of viewings and interpretations of the ceiling have been based. Essentially it refuses to acknowledge the role of the viewer or of time in the work of art. Instead, its authenticity as an act of painterly editing derives, or claims to derive, from Michelangelo—mostly, but with a little bit from the Council of Trent. The shuffle is interesting, acknowledging as it does a contemporary synthesis—an appeal to conventions in the present—while the rhetoric of the justification is an appeal to the past and to authorship (interestingly, curators use this latter term). In other words, the painting in the Sistine Chapel is an ongoing work which continues in the activities of Dott. Colalucci and his team.

The conclusion that the Sistine Chapel painting is an ongoing work may be felt provocative—or, at best, teasingly paradoxical. We normally prefer to think of the work and the conservation (or editing) as two separate things: the latter, we believe, is brought to bear on the former. The work is the quarry which, in all its details, the conservator and editor seek out. Such is Dott. Colalucci's aim, and we might assume it to be widely applicable. Leonardo's

Last Supper (1498) is, however, an immediate exception. Due to failures of adhesion, there have been extensive losses of pigment and priming across the whole surface.[6] Leonardo's work exists now only in fragments; the current restoration being undertaken by Pinin Brambilla Barcilon is intended to strip off the successive overpaintings to recover what remains of Leonardo's own work. Martin Kemp remarks that the resultant emphasis on 'the ravishing detail that will reveal the hand of the master' (p. 18) is often justified by close-up before-and-after photographs. The conservators' work is aimed at facilitating a form of reading the painting quite foreign to earlier generations of viewers and restorers, many of whom looked at it as a history painting and all of whom, according to Kemp, looked for the overall effect rather than to the fragment. In other words, modern conservational technology, justified by a rhetoric of authorial intent and authenticity, is re-presenting a Last Supper that is very much a product of the present. There can, therefore, be no 'real' Last Supper unrelated to the conservator. Kemp is quietly persuasive on the point:

The case of the Leonardo painting raises the question of how far the search for authorial traces should go. In the collection of the New South Wales State Library there is a painting of an important colonial figure, Elizabeth Macarthur, by an anonymous journeyman painter. Dating from about 1840, it had been overpainted in the Victorian period. Conservators removed the overpainting, thereby revealing a Regency-style bust but destroying a Victorian painting. The 'original' was uncovered, and it has welcome historical value; but the later work must also have had some which is now lost. Again, there is a large painting by Benjamin West from the 1790s, called Christ Showing the Little Children; it is in the Foundlings Hospital, London. It was subject to very early deterioration. West repainted it after ten years, and fifteen years later had once again to restore it. There are, thus, three versions of the work painted one on top of the other. All are authorial. Would there be grounds to remove one or both of the top layers? This would involve their destruction, but would reveal the original. Put another way: are there grounds, other than economic ones, not to remove the top layers?

Whichever way one answers the question, the need for the conservator and curator to assume an actively interpretative role is inescapable.[7]

Conservators then, rather like editors, do not occupy a neutral, ideologically innocent position that is legitimised by their scientific techniques. Post-structuralist theorists have repeatedly argued that we participate in what we know and that power inheres in the ways that we know. The application of this now-familiar argument is clear. Editors and conservators are not simply engaged in other-directed action. The artistic (or literary) work is not an object which of itself necessitates certain criteria in the restorative activity. One such criterion, authorial intention, is, according to Martin Kemp, too slippery a notion to be of use without redefinition in a way that deprives it of its capacity to encompass the wholeness or integrity of the work. Nevertheless, authorial agency remains of prime interest to most viewers and readers, and this factor can offer a justification for certain forms of editorial treatment of visual and printed artifacts where, at the local level, distinctions can often be made between traces that are probably authorial and those that are not. This justification is pragmatic rather than idealist: conservators and editors are interventionists, go-betweens, rather than scientifically disinterested technicians. In and through their activities, literary and artistic works continue to function—and function differently from before: editors and conservators broker workable solutions between documents and new readers, between artifact and new generations of viewers, on the basis of criteria which enjoy a currency and persuasive power in their day. Where this leaves the traditional editorial notion of the work (whose authorial text editors characteristically aim to isolate from the similar but variant texts produced by differently motivated copyists, printers and commercial publishers) is a question the remaining illustrations may help to clarify.

III

Here is Hogarth on the problem of change in the life of a painting:

Compare this with a more recent statement by John Brealey: Both Hogarth and Brealey defer to the artist's intention, yet both recognise that its recuperation is likely to be impossible.

The unavoidable conclusion is that total cleaning of a painting usually cannot restore the painting as it left the artist's hands even when it is completely intact. All it can do is reveal the present state of the paint beneath the grime. Thus, rightly to appreciate as viewers what we see in front of us at the art gallery, we have to recognise the agency of time in the life of artistic works and to appreciate the limitations which the passage of time imposes on the 'editing' of paintings. The viewer needs to know whether the painting has been totally cleaned, partially cleaned (this is where 'the surface of the original paint is not revealed, but remains covered by a thin layer of varnish' [Hedley, p. 163]), or, thirdly, nuance-cleaned (where differential amounts of cleaning are done in order to 'restore the relationship of values that, it is believed, would have existed in the original' [Hedley, p. 164]). Obviously both partial and nuance cleaning involve critical and aesthetic choices. Partial cleaning seeks to maintain the harmonising function of the first thin yellow layer of varnish and acknowledges its cultural function as a signifier of 'the age, the antique character' of the artifact (Hedley, p. 163). Nuance cleaning stresses the recovery—although it is, unavoidably, the renewed creation—of balance and unity in the painting.

Gerry Hedley, whose paper I have been quoting, stresses that where the original painting has undergone significant change (and scarcely any painting survives a century without physical damage) the artist's intention will not be recoverable in full. The best that can be done by the conservator, he argues, is the institution of 'newfound relativities to that intention and to time' (p. 164). The artist's agency was indispensable; but, as far as the work is concerned, that is not the end of the matter. The 'mortal body' of the painting, as one conservator calls it,[10] is always in a state of physical change; and at

crucial moments in its history it is, as it were, partially rewritten—re-published in a revised form—by its editor-conservators. Whatever approach is adopted the fundamental artifact is changed, taking on new meanings.

If artists are still alive when their work deteriorates then they may, like Benjamin West, act as their own conservators. Taking on this role creates new twists in the definition of the artistic work akin to those created for the literary work by the writer's returning for revision when, some years after publication, a second edition is called for. Take the case of Liz Magor, a well known artist-sculptor in Toronto who acknowledges the influence on her of Walter Benjamin's essay, 'The Work of Art in the Age of Mechanical Reproduction'. For Magor and many other artists in the 1970s the phenomenon of mechanical reproducibility tended to undermine commitment to the individual creation of the unique art-object in favour of transformation of preexisting cultural artifacts. Hence her description of one of her works, Time and Mrs Tiber (1976), which the National Gallery of Canada purchased in 1977:

Magor says she thinks of 'the life of a work as correlated to my own' (p. 11), but the Gallery (or 'museum', as those in the trade prefer to call them) had longer-term ideas. So when in 1987 three jars were found to contain botulism she suggested that she simply replace them, and she was fortified by the fact that the three had not been among those put up by Mrs Tiber but ones which she had originally prepared herself so that the dresser's shelves would be completely filled. This time she got assistance from an experienced hand at preserves, Mrs Coburn. Another of her works, Dorothy: A Resemblance (1980-81), in the same gallery, consists of four tables, each covered with small objects cast in lead: bottles, pears, slices of bread, books, small loaves, etc. Each object had been cast in moulds which she had made herself, but in rather primitive circumstances: the pears had been produced from one mould, the loaves from another, and so on. Trouble started when light fingers made off with some of the temptingly pocket-sized items on the tables. The Gallery contracted another artist to make new moulds from remaining examples on the table and so to cast replacements. However the replacements turned out to be not exactly the same as the objects the moulds were made from because of the superior equipment and technique which the contractor made use of. Magor's chagrin is evident: 'In spite of the fact that it is a piece made up of many parts, it was [originally] forged as a whole, and any replacement is unable to be part of that moment' (p. 10).

Magor's problem is the problem of all conservation, just as the problem of the author's revision is the problem of all editing: the later activity (assuming its presence is established) is unable to be part of the original moment, whether or not undertaken by the same person. Hershel Parker has demonstrated the problems for editors who would attempt to conjoin in a single reading text the results of radically separate acts of composition or revision; and many of us must have questioned the wisdom, even as they marvelled at the execution, of editions such as the Clarendon Tess of the D'Urbervilles which incorporates into the manuscript base-text some revisions made by Hardy more than twenty years after his original writing.[12]

This raises the question of textual authority: the right of the author to go on controlling the work's meanings. Editors have traditionally respected the author's right to do that. And the Canadian Art Gallery, in consulting Liz Magor about conservation, was doing much the same thing. But in both cases we are likely to confuse documentary ownership with textual authority. Having sold her pieces—her physical documents, as one may say—Magor no longer owned them. That is uncontentious, but one can go further: although she thinks of the life of her work as correlated to her own, its life goes on predominantly in the viewing and thinking of other people, including the Gallery's conservators. Over these readings—these 'textual' activities—she has little control and no authority. Magor's statements of her intentions and the contexts of those intentions in her life and other work will and should influence those textual activities: but only as part of viewers' efforts to understand and contextualise and conserve. Whether she should have the right to reconfigure or replace aspects of the physical artifacts is moot: she is acting as a conservator rather than artist if she does so.

The original moment cannot be recalled, as Claes Oldenburg also found when he replaced the pickle on top of his four-and-a-half feet Giant Hamburger of 1962. The piece is made of sailcloth, foam rubber and newspaper; the original pickle had been destroyed, but Oldenburg had offered to replace it for exhibition in 1967 in Ottawa. He made it in New York, and flew up with it to Ottawa, using it as a comfortable headrest on the way. When it was inspected at the Gallery, the paintwork was found to be significantly different to the original and more characteristic of his recent 1967 work than his earlier technique.[13] Both examples show that the personal and historical locatedness of a work is unavoidable and unreproducible even in works which, in their apparent mechanical reproducibility, embody their creator's rejection of the notion of the unique and inspired work of art.

IV

The moral is straightforward and surprising: whether they date from the twentieth century or the sixteenth, the artifactual remains become, at the moment of treatment, the site of conservatorial participation rather than the object of restoration. In the case of paintings the surface is partially re-made in a diachronic collaboration between artist and conservator which issues in a changed artifact. Conservatorial treatment, then, participates in the ongoing life of the work; and if the two coalesce for a time, they should not be thought of as completely discrete categories. If in addition, as suggested above in relation to Liz Magor, the viewer also plays a role in the work of art—at the level of textual rather than documentary activity—then a notion of the work as participatory begins to emerge.

One is immediately led to reconsider traditional conceptions of the work. The usual assumption has been that the viewer plays no role in the painting but only reacts to it. Put another way, the viewer may look at it, but is not part of it. However, this common-sensical appeal to the normal experience of museum art ignores not only the specialised viewing of the conservator at the point of treatment but also the historical fact that the very question of what constitutes a work of art rests squarely within the province of the viewer, including expert viewers. Take the case of J. M. W. Turner's paintings. When he died in 1851 he left his paintings to the British nation. This led to a legal battle about which of his works were finished paintings—'finished' being the term he used in the final form of his will. An inventory taken by the Keeper of the National Gallery in 1859 states that there were 100 finished pictures, 182 unfinished and many thousands of drawings and sketches. 'The wish expressed in the will was fairly strictly observed';[14] the remainder were consigned to the vault of the National Gallery, uncatalogued, and did not appear in the first catalogue raisonné of Turner's works which appeared in 1902.

But by the end of the nineteenth century, the definition of what was considered to be a Turner painting in the full sense was changing, influenced by the new taste for Impressionism and by a nationalist desire to compare late Turner favourably against the French Impressionists. Thus influenced, in the first decade of this century over seventy Turners from the vault were formally accessioned, including such famous ones as Norham Castle, Sunrise and Interior at Petworth. What are now considered towering masterpieces were not considered by Turner or by the National Gallery in the mid-nineteenth century to be works of art suitable for exhibition. With the benefit of Impressionism, curators and the viewing public decided otherwise. With the paintings transferred to the Tate Gallery, the process of accessioning continued, until by 1944 all of his unfinished paintings and his thousands of sketches—almost whatever his hands had touched—had been accepted.

The sense in which a work of art is completed only in its viewing, and

the contexts of that viewing, was clarified at the Canberra conference by a paper given by Christopher Saines who discussed the Queensland Art Gallery's Resurrection (?1560s) by Tintoretto, then on loan to the Australian National Gallery for its Rubens and the Italian Renaissance exhibition (1992). That exhibition was very consciously contextual in its educational strategy, aiming to demonstrate connections and influences over a couple of centuries. The finished painting in question has its own fascinating context within Tintoretto's œuvre and within the micro-history of its own production. Tintoretto did seven paintings of the Resurrection, and there are a further four workshop examples extant. The same studio prop, a large box representing Christ's tomb, appears in many of them. Analysis of the subsurface of this particular Resurrection has revealed that the composition was worked up on the canvas. In the course of the painting, Tintoretto must have realised that the theatricality of the Resurrection would be heightened if he brought the tomb to the very front of the picture plane, thus having the arising Christ appear to step right out of it into the viewer's space.

Perspectivally, the painting was intended to be seen high, as on a church altar, and one would have had to get down on the floor of the gallery to get anything like the originally intended effect. In other words, Tintoretto factored the viewer's position into his painting; he certainly never intended it to be hung in the Queensland and Australian National Art Galleries at eye-level.[15] A painting's size in relation to the viewer's size is important; indeed, trompe l'oeil effects depend on one's physical body and its expectations about space. One's engagement with a painting, that is to say, is not only intellectual, aesthetic and emotional: it is physical as well. Knowledge of it is partly corporeal. A work of art with a still unfolding mission, it comes into being with each new viewing. This is not to say that one's standing in front of it affects the artifact, except in microscopic ways. But, as a physical object, the painting is important—is a work of art—only in so far as it allows these personal encounters to keep on happening.

The necessary distinction here is between the painting as physical object (manifesting the agencies of artist, time and conservator) and as meaningful text (the agencies of viewers). The distinction, if accepted, has notable consequences. One is the provision of an answer to the iconoclastic argument that a work of art by a master which is subsequently exposed as a forgery has not in itself changed at all. Therefore, the argument goes, the notion of authenticity is a sham, and reverence of work 'by the hand of the artist' is a plot by auctioneers to keep up the market prices. However, this argument runs together the painting's existence as artifact with its capacity to produce meaning for the viewer. The latter existence (what I am calling 'text') feeds from contexts of understanding, including historical ones, in one's engagement with the painting. These contexts are ineradicably altered by the revealing

of the forgery: as a viewer of the forgery one had 'got it wrong'; one has to try again.

The second consequence is that, in the case of the literary work, 'text' (defined as before) requires the reader's involvement in raising meaning from what is ink on the page (i.e. 'document'). If seen as a combination of 'document' and 'text', works of literary (or painterly) art cannot be thought of as stable, or stabilisable, things. They function only in time and are never completed unless they become entirely ignored; and they have no unchanging existential anchor—whether in the author or in the historical moment of their creation or reading. Always subject to the processes of authorial reworking and revision prior to their publication or exhibition, they enter into a different kind of process in their public life.[16]

This line of argument is about the functioning of artistic and literary works. Concentrating on art as a phenomenon in a constant state of change offers an alternative view of the literary work to the traditional one as an abstraction hovering behind the extant documents (and, as is often said, 'witnessed' by them). Granted, our pragmatic agreements and conventions create the expectation that the work exists in its own right, statically and in isolate singularity. But the evidence suggests a constant state of textual activity on the part of creators, editor-conservators and reader-viewers. All are involved in the work: this is why conservation and scholarly editing must be understood as interpretative rather than scientific or technical activities. Some of this activity results in the alteration of the documents, the artifacts, from which future readings, future textual work, will be done by readers. Editors painstakingly re-configure and re-present a documentary form of the literary work; conservators must alter, even partially destroy, the artifact itself.

This last distinction between the dealings of editors and conservators with their artifactual materials might be thought to undo the line of argument about the work as participatory. However the distinction is not quite as it first appears. In part because of the persuasiveness of the editor's textual argument and in part because of the prestige of the press, the series, even at its crassest the amount of public moneys consumed in its preparation and publication, a new critical edition tends at least temporarily to eclipse other editions of the work—just as the newly conserved artistic artifact eclipses the old. Nevertheless, the significant difference remains that the pre-existing literary documents, unlike artistic artifacts, remain available for other editors to work on. In theory, if rarely in practice, multiple critical editions of the same work could be in circulation at any one time. Even if their editors agreed in their essential aim—the establishment of the text of final authorial intention—the editions would in practice be different in many details: the determining of intention is necessarily an interpretative matter, as is the postulation of readings from lost documentary states; and the deciphering of a difficult holograph depends on experience and skill. The various editors would be aiming, as it were, to release the text of the work from the documentary detritus accumulated in successive copyings and printings over the decades or centuries, to have the previously obscured text of the author's communication emerge as clearly as possible, to make their elusive textual quarry available in a reliable form for the first time. If editing has had an allure, this surely has been it.

Thus would the 'work' be figured, in actual editorial practice, as something over and apart from the editors and their activities even if (and to some extent because) no one of the editors will have perfectly achieved his or her aim. But what might not be clear to these various like-minded editors is that, as the comparison with paintings shows, they would be not so much 'releasing' the work as—rather like the conservators, though without dire effects on the existing artifacts—participating in its ongoing life, constituting its functioning, for the time being in an influential way.

I am entertaining, then, a phenomenological view of the work. Under this approach, principled disagreements about the modes in which editorial and conservatorial interventions should take place would remain as important as ever. But the view enforces the realisation that we would be disagreeing about our conventions for understanding the work, not something inherent in it—for there is no it, as work of art, independent of our understandings. On the other side of the coin, those efforts of understanding have an obligation to deal with what the physical artifacts, the documents, mutely testify to—and the documents, as documents, persist essentially independent of our understandings. In other words, our textual and editorial activity should be linked to, and seek to understand and contextualise, prior documentary activity: oral transmission aside, that is the basis of our access to the past. Accordingly a phenomenological view would abandon any belief in the work as an ideal thing and instead acknowledge (for possible bibliographical analysis) the existence of documents from the past, what people have done with

them in the composition, production and reading of them, and the historical locatedness of those acts. In that terrain editing would make its home. Its services would be found just as necessary, but it would figure itself differently from before.

Notes