Why Connoisseurship Matters
David Freedberg
In 1998, in an article intended to accompany his great mmma on Flemish Art and Architecture in the Pelican History of Art series, Hans Vlieghe commented on 'the significant increase in the number of new contributions on seventeenth-century Flemish art since I960'.1 What he did not note, or course, was that he himself was responsible for much of this increase. He has ranged more widely in the field and refined it more thoroughly than any other scholar in the area. Thirty-one years earlier, in 1967, he wrote:
'Ons beekl van de onrwikkeling tier Antwerpse compositiesehilderkunst in de zeventiende eeuw is, behafve voor war her oeuvre van Rubens, Van Dyck en Jordacns aangaat, weini^; diiidelijk. Voor een belangrijk gedeelte is dir te wijten aan he.t feir dar veel van de schilderijen die door Antwerpse schilders van het tweede plan zowat in de sehadtiw van de grote kunstcnaarstrits van de Scheldestad werden uitgevoerd nook gepub-liceerd zijn gewotden, ja zich in heel wat gevallen under fuutieve benamingen of als anon-ieme stnkken in minder bekende kerken, parti-i ulii-i t- vetzamelingen of moeiiijk toegankelijke museumdepots bevinden.'^
In fact, Vlieghe had already begun remedying this situation, and from the time of the
epochmaking article in which he made this statement, he set out to clarify the oeuvres of the lesser-known Antwerp painters and bring them out of the shadows. He rectified the mistakes and filled the lacunae with energy. He scoured rhe remote churches, penetrated the private collections, and found his way into the obscure depots in order to reattribute the works and refine or consolidate the oeuvres of the then lesser known Flemish masters. They were 'lesser-known' then, but now we know them with vastly greater refinement than, when Vlieghe first began publishing his articles, beginning with his investigations of David Teniers II and III precisely in 1960.J
'Het omrstellend gebrek aan behoorlijke infor-matie aangaande de Vlaamse schiiders van grote religieuzt en historische taferelen, die in de XVIIe eeuw zowat op het rweede plan hebben gewerkt, is een merkwaaidig reit',
he noted in the second of the articles he produced in the annus mirabilu that was 1967.^ This was the article in which he almost single-handedly laid the basis for the creation of the oeuvre of an artist until then almost entirely known from prints and documents, Cornells Schut. Before then, as Horst
WHY CONNOISSEURSHIP MATTERS
Gerson noted in the Pelican History of Art volume which Vlieghe's summa replaced, Schut was only known for his engravings, while his paintings were 'concealed amongst a host of uncertain attributions'.-1 Just as he would later do in the case of many other painters, Vlieghe set out a number of secure attributions, which then formed, the basis for the now clearly-defined corpus. Ir was precisely in the area of hisrory painting — once, of course the most important of the genres, but now most neglected of all — that Vlieghe made his most significant contributions. He did so thanks to his incomparable combination of documentary and visual skills - just what Berenson, in his classic essay on the task of the connoisseur demanded6 - as well, of course, as his keen sense of who the important figures outside the great trio of Rubens van Dyck and Jor-daens actually were. Our understanding of the visual culture that produced these rhree has been immeasurably enhanced by Vlieghe's own work. In this sense connois-seurship has actually enabled the study of what is too often tegarded as the modern replacement of traditional art history, namely that of 'visual culture', -a discipline that pretends to be new but in fact rests entirely on some of the most characteristic skills and interests of art histoty.
Indeed, already in the first of the benchmark articles from just cited, Vlieghe made fundamental additions and refinements to the oeuvres of Jan van Boeckhorst, Theodor Boeyermans, Erasmus Quellinus, Cornells Schut, Pierer Thijs, Theodoor van Thulden, and the neglected history paintings of Cornel is de Vos, Many of these are names to which Vlieghe often returned, frequently with trenchant effect, as in his notable resolution of the problem of the Cardiff cartoons
once so optimistically attributed to Rubens by giving them decisively to Jan van Boeckhorst,8 his fundamental and still suggestive article, on the Roman years of Cornells Schut,^ and the rich work on Erasmus Quellinus, Rubens's 'primer official'.10 It is a measure of the fruitfulness of Vlieghe's work that all of these contributions led to fully-fledged monographs by others on these artists — including the important 1990 exhibitions on Jan van Boeckhorst,11 Katlijne van der Stighelen's outstanding monograph on Cornells de Vos,12 and the thorough examination of the career and work of Cornells Schut by Gertrude Wilmers,1-1 ro menrion only a few. On the other hand, when it came to Erasmus Quellinus, Vlieghe pointedly expanded and corrected the picture given by Jean-Pierre de Bruyn in his monograph on that painter.14 Typical is his comment on the Homage to Moses after the Passage through the Red Sea. After pointing out that De Bruyn only mentioned three of the four oil sketches in the cycle, Vlieghe promptly locates the fourth sketch, and then adds laconically and devastatingly 'Deze schilderijen zljn echter typisch vroeg werk van Jan van den Hoecke',15 How few of us could immediately identify a Jan van den Hoecke, let alone be sure of his typical early works! So too in many other cases, far too numerous to mention. Throughout the sixties and seventies, Vlieghe cleared the ground — and the air — to give us sharp profiles of many artists, ranging from Gas-par de Crayer, on which he wrote his fundamental monograph,16 right through to lesser figures such as Arnout Vinckenborch and Artus Wolffort, whose work would have been entirely lost to the history of art had he not identified and thereby resuscitated them.1^
Who therefore, looking at the work of Hans Vlieghe could say that connoisseur-ship does not matter? Or that is somehow a peripheral task? It lies at the very core of the history of art.
In 1719 Jonathan Richardson published his Two Discourses", in which he set out the aims and principles of connoisseurship.[S These were its foundational arguments, though of course its practice already had a long history. For Richardson, as, probably, for most practitioners, connoisseurship had three basic aims: firstly the making of judgment of quality ('the Goodness of a Picture' as Richardson put it), secondly the assignment of hands, and thirdly the distinguishing of originals from copies (including, of course, the identification of forgeries).1^ In the second of the Two Discourses Richardson set out why connoisseurship was a suitable task for gentlemen20 — a designation which has come to cause much trouble for the subsequent reputation of connoisseurship. Now the present discussion will not chiefly be concerned with judgements of quality, if at all. It will be confined to those aspects of connoisseurship in which Vlieghe excelled, namely the establishment of the oeuvres of masters, and, pari passti, the distinctions between originals and copies, and the identification of forgeries. These aspects of connoisseurship do not, in and of themselves, have much to do with the determination of the artistic worth of a painting or sculpture (vague though this pursuit may be), but offer the possibility of arriving at falsifiable and therefore secure conclusions.
It is easy enough, prima facie, to give examples of why connoisseurship matters, of where it has been satisfactorily used to establish a corpus of works and where it has usefully identified copies and fakes.21 These activities have a kind of self-evident,
(though sometimes inflated) worth. It is even easier to say why connoisseurship does not matter, or where it is flawed or corrupt. Many of us, enjoying a painting, would agree that it really does not matter from an aesthetic point of view whether the Dresden Venus was painted by Giorgione or by Titian, the Polish Rider by Rembrandt or Drost, the Woman in the Red Hat in Washington by Vermeer or by a nineteenth century faker. But if we feel — or think we feel - so secure about the aesthetic merits of a painting that it does not matter who actually painted it, then it is important also to consider the strong counter-arguments in Nelson Goodman's Languages of Art. In this book — insufficiently read by art historians — Goodman argued powerfully that simple knowledge of the fact that a work is an original (and mutatis mutandis not a fake, or by someone else) is as critical a factor as any other in our esthetic judgment about it,22 This formed part of Goodman's complex and brilliantly argued view that the cognitive is an essential part of the esthetic.
One ol the easiest arguments to make against connoisseurship, to say why it does not matter, is to point to the social and financial pressures on judgment, particularly judgments of attribution. There is no shortage of examples of the way in which financial pressure and inducement influence the upward attribution of a work. There are too many of such cases, and rhey are often shaming, Markets — galleries, dealers, auction houses - exercise undue pressures. They induce experts to make judgments
- both of quality and of attribution - that suit their own pockets rather than the purposes of disinterested scholarship.
This weakness of connoisseurship
— though I do not believe it is an inherent one — is closely related, if only because of its
WHY CONNOISSEURSHIP MATTERS
social dimension, to the several ways in which connoisseurship may be fetishized. The totem ization of the superior eye is a familiar enough phenomenon, and so too is the enhanced social standing sometimes achieved through it. Bernard Berenson provides the classic example. He sorted pictures out, spoke eloquently about them, was thoroughly corrupted by dealers, and operated from a chic villa in the Tuscan hills, which both enhanced his own social standing and that of those who visited it. But few, if any, would now admit that connoisseurship should matter because it makes one a better gentleman, as Jonathan Richardson thought.2^ Certainly most of us can agree that the skills of connoisseurship can make one a better art histotian, as Hans Vlieghe has consistently shown throughout his career (he would have been a perfect gentleman anyway).
But ler us return to the second and third criteria for connoisseurship, namely the ability to make judgments about authorship (that is, about attribution and the ability to recognize hands), and to distinguish true from false.
A lot of ink - quite a lot of good ink in fact - has been devoted to these. The Morel-lian method of making attributions on the basis of those features of a painting that are unconscious rather than conscious — fingernails, toes, tearducts, earlobes and so on — has actually yielded many good results over the years. Ac the same time, the dangers and the pitfalls or attribution by this method are obvious, since earlobes and tearducts are arguably not always unconscious; and wishful thinking so often does come into the matter — the wish to assign a name, to make the work be by a superior rather than by an inferior master, as well as the almost equally strong desire to down-
grade the work, to take it off its pedestal, even to declare it a fake. This latter seems so often to be an easier move to make than a more positive one.
Here I want to refer to four aspects of connoisseurship which, it seems to me, serve to restote to it some of the disciplinary esteem it deserves. The first two have received some attention already; the third has been much discussed, but nevet rigorously, and never in the new context I present here; and the fourth, though obvious, has received practically no discussion — at least no overt and systematic discussion — at all. The first relates to the epistemological status of connoisseurship (though it also reveals its disciplinary relations with other fields); the second relates to the interdisciplinary nature of connoisseurship (though it also has to do with its epistemological status, in obvious ways); the third, relating to intuition, places connoisseurship in its neuro-scientific context; rhe fourth, relating ro trust, examines the dimension of connoisseurship that is both social and cognitive, and thus falls into an equally new field, that of social-cognitive theory. I shall discuss each of these areas in the briefest terms, not pretending to thoroughness (anzif), but simply as pointers to the rich possibilities for reconceiving connoisseurship as a core discipline in the humanities.
Ir is easy — and has become very fashionable - to criticize the very practice of connoisseurship. It has had few strong and eloquent defenders in recent years; but there has been one exceptional one. He comes, at first sight, from an unexpected quarter. Carlo Ginzburg's vigorous and original defenses of connoisseurship have pointed to both its historical and its epistemological significance. As the teal founder of the school of what came to be called microhis-
tory (and from which he later distanced himself), Ginzburg showed how to build a large and fascinating narrative, a picture let us call it, on the basis of a careful study of particular and - this is significant - highly individual and peculiarly distinctive documents. And for him, as an historian, the importance of establishing the veracity and authenticity of documents has always been crucial.
In his brilliant article Spie, translated in 1989 as 'Clues: Roots of an Evidential Paradigm', Ginzburg showed how the practice of connoisseurship, and in particular the Morellian connoisseurship on which we all depend (even if only tacitly), forms part of the emergence in the nineteenth century of a central epistemological model in the humanities. As Ginzburg correctly noted, the Morellian method relies not so much on the analysis of the conspicuous and conscious characteristics of a picture' (which are of course the easiest to imitate), but on 'the most trivial details', the unconscious and unwitting ones, one could say, 'that would have been least influenced by the mannerisms of the artist's school'25. These are the tell-tale signs, or clues, as Ginzburg very deliberately called them, by which the identity of the artist are established. In setting out the contemporary historical connections between Morelli's approach, the procedures of Sherlock Holmes and Freud's Interpretation of Dreams, Ginzburg made it clear that connoisseurship was the emblematic discipline of that particular paradigm of scientific investigation that depended on the analysis of unconscious symptoms, and inadvertent or involuntary, rather than conventional gestures as an indicator of the identity of the whole. Though this epistemological model reached its apogee at the end of the nineteenth century, it had its
roots in a much older paradigm, that of the study of signs, what we would now call semiotics. In this way, connoisseurship took its place amongst all those sciences that served to establish the individuality of phenomena (or in this case of persons) through the distinctive traces they left behind. It was above all a conjectural, venatic discipline, in that it depended on tell-tale signs, and had to do with what Ginzburg called low intuition, a kind of basic intuition about the identity of things. It is predicated on judgments that are swift and apparently immediate and of a kind that experts share with ordinary folk and with non-experts.
About a decade later, in a short but important article in The Art Bulletin on the limitations of interdisciplinarity, Ginzburg offered a pointed defense of connoisseurship, and set out how 'rather than being opposed to an interdisciplinary or multidis-ciplmary approach', 'in most cases {it acta-ally} implied such an approach'.- He insisted upon the way in which the connoisseurship of objects such, say, as a group of nature drawings recently firmly attributed to Martin Schongauer, was not in fact simply a matter of intuition, but rather relied, in the end, on a large number of experts in other disciplines - from watermark and paper experts to botanists, archivists, paleographers, and so on. He then showed that Fritz Koreny's attribution of a drawing in the Getty Museum27 not only had obvious implications for the matter of authorship, but much larger implications as well, such as a much earlier date than usually allowed for the beginning of what might be called 'the scientific depiction of nature in Germany'.28 In other words, Ginzburg claimed that connoisseurship was not just, 'simply', interdisciplinary, but that it was through this very inter-
WHY CONNOISSEURSHIP MATTERS
disciplinarity that it offered the possibility of engaging the much wider world in which works of art are set. In his view, connoisseurship-5 could tell us not only about authorship, but about the world too. It has a 'cognitive richness', as Ginzburg put it, of which even its practitioners may be unaware.
It is not difficult to agree with Ginzburg's claim that conuoisseurship has become an interdisciplinary task (rather than, as it may seem at first sighr, a narrow and insular one) and the role of new instrumentation and techniques in the scientific confirmation of the intuitions and judgments of connoisseurship has become widespread and widely accepted.1'0 But the pitfalls remain, especially that of the inevitable subjectivity in the reading even of 'scientific' data; and it is through one of the most striking of them that we can attain a still clearer sense of why connoisseurship matters. Let us look at some of the problems associated with the ways in which we reach — or jump to — conclusions about originals and fakes.
Is rhe Apollo Saurochtonos recently acquired by the Cleveland Museum of Art really Praxiteles's original? Or is it a fake? To pronounce either way carries great risks with it, just as, say, in the case of a Cupid recently (1996) claimed to be by Michelangelo in New York,31 If you arc wrong you get egg on your face. But there are less high stake cases roo which are equally problematic, and which we frequently encounter in our daily work. There are hundreds of examples in Rubens alone. Vlieghe's connoisseurship has almost always been of a much less hazardous and more secure kind than the optimistic attributions at the upper end of the Flemish field. Even he, however, would probably admit that one of
the most striking features of the way we think as connoisseurs is the role played by intuition. We are all familiar with the experience of recognizing in a flash who a picture is by; and of recognizing a work as a copy, or a fake. We often have the impression that we would actually be better connoisseurs if we stayed with intuition entirely. While there is a substantial literature on the ways in which intuition leads to suboptimal decisions,32 connoisseurs have learned to lend it a certain credibility, to endow it with a sense of trust - even though we may be unable to set out exactly what lies at the basis of our decisions. The fact that intuitions in connoisseurship are frequently as good or better than judgments derived from deliberation1^ has recently received extended discussion in Malcolm Gladwell's popular book entitled Blink! where a number of the new scientific approaches to the problem of intuition are engagingly set out.'4 Indeed, as an introduction to the whole problem of problem of instant judgments, the flash decisions on which our intuitions are based, Glad well began with the example of Federico Zeri's instant decision that the now notorious Kouros acquired by the Getty in 1983 was a fake. Not only Zeri but a number of other distinguished arr historians immediately thought this alleged work of the 6th century BC was a fake, against the long-deliberated judgments — and the scientific reporrs — of the experts hired by the Getty. ^ The snap-decisions of the best connoisseurs often turn out to be the correct ones, From Berenson through Max J. Friedländer and onwards, many of them have been very aware of the role of intuition in the judgements they make about pictures, and many — including Berenson and Friedländer — have tried to discuss if;- ■ But what actually lies behind
these intuitive decisions, in which we are not at all conscious of what goes into them? This is a complicated issue, but we now are in a better position than the old theorists of connoisseurship to describe the subjective experience that is associated with the use of knowledge gained through implicit learning, 'the acquisition of knowledge that takes place largely independently of conscious attempts to learn and largely in the absence of explicit knowledge about what was acquired'.3'' How, in the first place, are we to speak of that subjective experience, beyond calling it simply 'intuition'?
A large quantity of the most interesting neuroscientific material is now available on exactly the mental processes involved in such intuitive decisions that turn out to be correct.oS 'They are fast and take into account non-consciously generated information, gathered from experience, about the probabilistic structure of the cues and variables relevant to ones judgments, decisions, and behavior'.^ Moreover, they are based, in the first instance, on the kind of editing out of small-scale but indicative features, such as suggested by Morelli (and futilely critiqued by Richard Wollheim);40 upon which judgments of the authenticity, or rather of the individuality of the whole are made (as in the case of Zeri's Morellian assessment of the fingers of the Getty Kouros). In making assessments of others, our unconscious finds patterns in situations and behavior based on very narrow slices of experience; and so too in the case of pictures. We may have built up years of acquired knowledge about a subject, but in fact when faced with a picture, say, we unconsciously edit out the particular feature^) that provide us with the key to authorship, or, more precisely, to what we instantly know to be characteristic of the
author's hand. We find evidence in the narrowest slivers of experience, which the best connoisseurs know how to choose well and in seemingly spontaneous ways. Just as in bird-watching say, you know precisely what details to look fot that are most indicative, without engaging in long and deliberate analysis. No wonder that so many of the late seventeenth and early eighteenth century connoisseurs, just like military strategists, spoke of the coup d'oeii, or better yet I'effet an premier coup d'oeil, as the art theorists more precisely called it ''Ll You know what is before you at a glance. This critical part of rapid cognition, whereby you instantly select out a particularly indicative derail or characteristic, is what is now called thin-slicing, a phenomenon which has to be studied in conjunction with the crucial role of the ventromedial prefrontal cortex in making intuitive judgments, as so well described by Antonio Damasio in his famous description of the Iowa blue and red card gambling test.42 Even before we arrive at articulate conclusions our bodies often register their anticipation (in the case of the card test by sweating palms, slightly raised heart rates, and so on); and we reject unacceptable hypotheses well before we are aware of doing so. It turns out that snap judgments ate not only quick, they rely also on the thinnest slices of what might possibly be long-accumulated experience. We also have quite substantial information, thanks to the remarkable experiments by Benjamin Libet and others, about the ways in which swift judgments - which depend on the rapid but unconscious synthesis of swiftly gathered data - actually precede our conscious judgments by a lapse of milliseconds, and we know where in the brain these plottable intuitions occur, beginning with the basal ganglia and - not surprisingly — that part of
WHY CONNOISSEURSHIP MATTERS
the brain so clearly involved in memory, the hippocampus. ■ Both play a critical role in the phenomenon of implicit learning, that form of learning, as noted above, that occurs outside conscious awareness.
In this way a problem in folk psychology, as described by Ginzburg in Clues, joins with the epistemology of connoisseurship and the latest findings of neurosdence. Connoisseurship does indeed occupy a place at the center of the disciplines — and not just the center of the discipline of art history — not only because it now uses a multiplicity of different tools, but because its essential epistemology suggests an exemplary crossover between the new neutosciences and traditional historical ones.
But there is more. Let us turn from this ail-too brief incursion into the relevance of the cognitive neurosciences to social-cognitive theory. If there is one area in the problematic of connoisseurship that has never adequately been discussed, it is that of the role of trust and authority. What, we need to ask, are the gtounds for epistemological authority in this field? The question touches on the older field of sociology and the newer one of cognitive philosophy. We can all agree that in making attributions we rely not only on our own judgment, but also, to a greater or lesser degree, on the opinion of the best possible authority in the field. But on what bases do we decide to trust an authority? How do we determine the criteria for deference? The easy answer is to say that we rely on the most impartial judge of paintings, the one who is least likely to be swayed by market or social pressures, the one who has that indefinable quality, "the best eye', as we so often like to say. But how do we decide all this? It is dear that the issue of whom to defer to is central; but what are the bases of our decisions about whom we decide to trust? Moreover, as Diego
Gambetta put it at the end of his 1985 anthology on the subject: 'how, indeed, can we trust trust?'44 To some extent, of course, we trust ourselves, our own judgments, but certainly not only ourselves, and very definitely not when it comes to the judgments about paintings and sculptures about which we are nor ourselves experts. We tend, for the most part, to follow the judgments of those who are both acknowledged as experts and whom we ourselves can acknowledge as such. But obviously these two decisions do not always coincide. We all know of any number of cases in which x is supposed to be the best judge in a particular field, but where, for one reason or another - perhaps because we know that x's judgement is not always entirely impartial or uncorrupted — y, despite his lesser fame and status, is in fact a much more reliable judge. How, in fact, do we arrive at such decisions, and how can we know that they are to be trusted, as Gambetta put it?
Obviously, for example, we trust those wrirers and experts who have proved their reliability by rheir publications, and we especially trust those whom others we trusr trust. Much more than in most other disciplines, however, we are aware of the degree to which the publications of connoisseurs may be tainted by external pressures. Too often authority in the field is established on the basis of evanescent matters like public acclaim, success in networking, and so on. The basis for such authority can turn out to be spurious or corrupr. But this is precisely why connoisseurship does matter. More than in other disciplines we come to discover that we need to look past the ephemerality of fame, worldly prestige, and all the other spurious elements on which authority and trust are established.
For some time now, scholars in the history of science have been concerned with
the ways in which trust plays a defining if not imperative role in the constitution of very kind of knowledge. The landmark book in this domain is Steven Shapin's remarkable A Social History of Truth, Civility and Science in Seventeenth Century England (1994). As its subtitle implies, Shapin sets out at some length the relations between gentlemanly practices and scientific ones in seventeenth century England. He shows quite precisely how scientific conclusions
— especially the ones that turned out to have the most lasting validity — were precisely rhose which gained credence as a result of mutual trust amongst those who had gained a repuration for reliability, relevance and appropriate scientific procedures. Now it's true that connoisseurship may seem to be vitiated by the link between gentlemanli-ness and science. But the overarching argument of Shapin's book is to show how truths are socially constructed; and it is in this context that we need to view connoisseurship too, realizing that its deficiencies in practice do not necessarily entail a fatal deficiency in principle. Shapin's argument shows precisely how important this link between science, trust, and civility can be in constituting knowledge. Noting that 'the identification of trustworthy agents is necessary to the constitution of any body of knowledge',45 he examines the many ways in which trust, civility, honor and integrity
— in other words the very codes and conventions of gentlemanly conduct — were rhe means by which problems of credibility in seventeenth century science were resolved. Does this not sound like the ideal connoisseur, the very embodiment of qualities that we find in Hans Vlieghe himself? We can all think of instances where the link between gentlemanliness and science has been abused; but Shapin clearly demon-
strates how the qualities that built up epis-temological trust in seventeenth century England were subsumed under the codes and conventions of genteel and civil behavior. Could we not imagine a stare of affairs in which the codes and conventions on the basis of which trust is established are more rigorously adhered to, and developed in such a way as limit the pressures to which so many connoisseurs are subject?
It is certainly the case that even the extraordinary scientific achievements of Robert Boyle, for example, to say nothing of Galileo,4^ had their own particular moral economy - although in their cases it was largely beneficial, despite what we would now call its downside. They worked in a context of trust, and their results and proofs were were more readily accepted because of that context, in which the constituent elements of trust were clearly defined. These elements were embedded in the procedures of science and followed a relarively strict and certainly well-defined epistemological decorum. Following Shapin's studies of the gentlemanly context of seventeenth century Academies and scientific procedures, more and more work has been done to show how courrly practices and the associated pursuit of elegance were essential aspects of the behavior of seventeenth century scientists. Shapin argued that the codes of genteel conduct which he described for seventeenth-century science formed and still form, an important basis for securing knowledge about the natural world. Even in what we would regard as the incontrovertible results of someone like Robert Boyle, we find a clear relarionship, just as in connoisseurship, between gentility and veracity 4
In the end it is not a question of the goodness ot badness of the moral economy of connoisseurship. What connoisseurs hip
WHY CONNOISSEURSHIP MATTERS
needs is a more positive epistemoiogical grounding. More aware of the ways in which trust may be constituted, we could strive not to trust an authority just because she has the trappings of civility and gentility - since we all know that such trappings are purchased at a price, and are open to be bought. What can be done is to institute an awareness, as far as possible, of the ways in which the assessment of testimony is made on the basis of its multiplicity, plausibility, directness, knowledgeability and so on. Such qualities could then be seen as the consequential systematization of prudential wisdom circulating in a culture, gentlemanly or otherwise. It should indeed be the criterion of gentlemanliness, or to put it in less gendered terms, of the essential decorum of scholarship, for the practitioner to have the genuine and proven skills to sort out both tellable and unreliable testimony, and reliable and unreliable judgments. Epistemoiogical decorum, to use Shapin's entirely appropriate phrase, thus proves to be a central element of good connoisseur-ship, and the key to its success or failure. There is of course a moral dimension to all this; but what is critical is that in thinking of connoisseurship in this way, we also learn to understand that it is not so different from the commonest ways in which truths are constituted — even in the best of rhe sciences — and of how often social facrs underlie what we would call objective science. The fact we that often can't trust the most well-known connoisseurs, or that their motives are transparent, does not mean that the effort is flawed in and of itself — just that it is more open to distrust than many other fields.
In an important recent paper, entitled
'What does it mean to trust in episiemic authority?' Gloria Origgi further explores the prob-
lem of sociological and moral theories of the trust. ^8 Taking issue with the non-cognitive view of trust, that is the view that we are disposed to be trustful of others independently of our beliefs or expectations about their trustworthiness, she brings cognitive theory to bear on the whole subject, And she goes much further than Shapin in interrogating the grounds on which we choose to defer to the authority of another. She explores what she calls 'the cognitive division of labor', in other words, the ways in which, just as in connoisseurship, we acknowledge that we cannot investigate everything ourselves, and that in very many cases we actually need to trust the authoriry of others. In the first instance we have to believe that what the speaker has or is going to say is relevant to us, so that we can even bother to interprer whar she says. Trust is a fundamental element of social communication, and 'the presumption of relevance that accompanies every acr of intentional communication is what grounds our trust in others'.^
We also have to believe that the speaker, that is the expert, does not say what she believes to be false.5U In other words, there has to be a spirit of cooperation, for we operate in what Dan Sperber and Deirdre Wilson in their work on Relevance Theory describe as a 'mutual cognitive environment'.51 This is 'set of hypotheses and representations that we activate in our mind when we communicate in order to understand each other'.52 We have to be able to believe that we ate, as we say, on the same page, or — to mix metaphors — in the same boat together. If we did not feel this about our authority, we would not be able to defer to her at all.
But there are other critical factors in connoisseurship too, no less - and perhaps even more - than in the case of other acriv-
ities where trust is crucial. Trust depends on both the competence and the benevolence of the trustee. Competence is an obvious requirement; but it is benevolence that turns out to be critical (you can be incompetent and still hit the mark, but malevolence or, indeed, any form of subversion of benevolence can only do mischief). We can either track your past record in a particular domain (and grant you authority on the basis of your earned reputation in this domain) or we can grant you authority on the clear basis of your better epistemic position (eg. first-hand examination of works, expert use of any one of a multitude of the new techniques now applied to works of art, from dendrochronology to spectrographic analysis, from infrared reflectography to autoradiography, and so on and so forth). But we would instantly walk away from someone we might be expected to trust if we felt that she were infringing the bases of our common commitments — to the field, to the objective analysis of authorship, and so on.
There are several complications here, however. Above all, we know too well from connoisseurship that we cannot count on the benevolence of many connoisseurs, especially when they use their greater or lesser skills in the service of the market or for social advancement. Their commitments may in fact differ from ours. So, as Origgi points out, we cannot appeal only to our expectations of other people's motivations to commit to cooperation. So then what are we to do? Origgi correctly notes that our commitment to trust is not only cognitive, based on the degree of our beliefs about the future decisions of the trusted. Our trust in others also involves what the literature in the field would call the 'motivational' dimension, that which depends on out
deeper emotional, moral or cultural pre-commitments. Now Oili Lagerspetz pertinently observed that 'trust is not the fact that one, after calculating the odds, feels no risk: it is feeling no risk without calculating the odds'.''-1 This is precisely what is at stake. Trust is fundamentally — whether we like it or not — a form of risk-taking, however low we may assess this risk to be (and obviously, in the case of trust, the risk will necessarily be felt to be low — though not as low, clearly, as in the case of security, safety and certainty). In many circumstances we can't actually say whether someone is really reliable or not, since our ignorance or uncertainty is great. We just have to trust them on 'motivational' grounds. Origgi is skeptical about this form of trust, since it seems to be based so much our feelings, our emotional sense, so to speak, about the person concerned, or about their views. She does not like the idea of making (or simply having) an emotional commitment to the person (or the view) to whom we grant authority. But I happen to believe — rather along the same lines as Nelson Goodman — that the emotions also function cognitively. They often help us arrive at more just assessments, even if it is as the result of weighing emotional commitments — or ideological ones — against the facts of pictures.
But there is also, I think, a yet more significant epistemological turn when it comes to the role of trust and deference in connoisseurship, This may in fact provide a key example of the how the riskiness of endeavor can actually facilitate and enhance its epistemological status. As Gambetta noted: 'Being wrong is an inevitable part of the wager, of the learning process strung between success and disappointment, where only if we are prepared to endure the latter can we hope to enjoy the former. Asking
WHY CONNOISSEURSHIP MATTERS
too little of trust is just as ill-advised as asking too much.' 55
In other words, we need to accept both risk and vulnerability when it comes to the decisions we make. It is precisely this that is at stake in the issue of trust, and in which it significance lies. 'We trust our interlocutor in her willingness to share a mutually relevant cognitive environment, that is to build a common ground that maximizes understanding and favors the emergence of new, relevant thoughts'.56 It is true that we can then always check the proposal and reject it; but nevertheless, 'trusting in the relevance of what people say entails the cognitive vulnerability that we accept in order to activate in our mind new thoughts and hypotheses we then share with our interlocutors'.57 And so begins the chain reaction of hypotheses on which good science is always built. It always involves risk. This we have to accept. What we do not have to accept is the notion that connoisseurship is irreparably impaired, irreparably fraught with danger because the bases of our trust are incapable of having clear epistemological grounds.
In examining the work of Hans Vlieghe it would be hard to deny the importance of connoisseurship or why it matters. From the moment I met Hans — many years ago, in the first Rubenianum, in the old Museum Smit-van Gelder, where we passed many happy hours — I found in him a whole series of those qualities which I have described in this paper. It is not always easy to say what makes a good connoisseur, but in him it was instantly clear. Aside from proven skill and the critical knowledge of the documents, to be engaged at the appropriate moments, it had to do with exemplary impartiality, a knowledge of both high and low, indifference to worldly fame, and a tare detachment from the market. But most of all it had to do with the trust which his work has always inspired, and the consistent deference his authority commands. Both of these are built on precisely the qualities set out in the last section of this paper — high competence, benevolence, and above all the codes of civility and sheer gentlemanliness to which I here hope to have given some small homage.
NOTES
1 Hans Vlieghe, 'Flemish arc, does it really exist?', Simiolits. XVT, 199«, p- 187.
2 Hans Vlieghe, 'Nieuwe tťtesčhrijvingea aar« Antwerpsc Schilders uic de Zeventiende Eeuw', Gentse Bijdragen toi du Kunstgeschieden is, XX, i967, p. 159.
3 Hans Vlieghe, 'David Terriers II cn David Teniers III patroonschilders voor dc tapijtwcvcrijcn', Arte* Textiles, V, 1959-1.960, pp. 78-102, followed by a series of articles throughout the 1960s on both of these artists,
4 Hans Vlieghe, 'Nieuwe toescrmjvingeii ami Cornells Schur', jaarboek. Koninkiijk Museum voor Schont Kumten Antwerpen, 1967, p. 187.
5 Hurst Gerson and E.H. ter Kuile, Art and Architecture in Belgium 1600-WOO (Pelican History of Art), Harmondsworth, I960, p. 141.
6 Bernard Berenspn, 'Rudiments of Connoisseurship', in: The Study and Criticism of Italian Art, 2nd scries, London, 1902 [reprinted separately as Rudiments of Connoisseurship, New York, 1962].
7 Hans VlicgllC, 'Nieuwe toeschrijvingen aan Cornells Schuť, Joarboek Koninkiijk Museum voor Schone Kumten Aitt-
Wúrpétí,l967, pp. 187-198.
8 Hans Vlieghe, 'The Identity of the Painter of the Cardiff Cartoons: A Proposal', Burlington Magazin», CXXV, 1983, pp. 350—356 and Hans Vlieghe. 'The Cardiff Cartoons: Boeckhorst after all', Burlington Magazine, CXXIX, 1987, pp. 593-599.
9 Huns Vlieghe, "/.u den Römischen Jahren des Malers Cornells Schuť, Mitteilungen des Kumthistorischerl Instituts in Florenz, XV, 1971, pp. 207-21R.
10 Hans Vlieghe, 'Erasmus Quellinus and Rubens's Studio Practice'. The Burlington Magazine CX1X, 1977, pp. 636-643 and Hans Vlieghe, 'Een wazige kijk op Erasmus II Quellin. Bedenkingen bij recent onder^oek', Jaarhoek Koninkiijk Museum voor Schone Künsten Antwerpen, 1993, pp. 287-311.
11 [Cat. cxh,]Jí<>2 Boeckhorst, Antwerpen, Rubenshuis / Münster, Westfälisches Landesmuse 11m, 1990.
12 Karlijne van der Stighelen, De portreiten van Cornells de Vos (1384/5—1651J. Een kritische calalogus (Verhandelingen van de Koninklijke Academie voor Wetenschappen. heileren en Schorn Künsten van Belgie, Klasse der Schone Künsten, LH, 1990, Nr. 51), Brüssel, 1990.
13 Ida Gertrude Wilmers, Cornells Sckut (1596-1655). Flemish Painter of the High Baroque (Pictura Nova, I), Turnhour, 1 996.
14 Jean-Pierre de Bruyíi, Erasmus IT Quellinus. De SchiUerijen met Catalogue Raisonné (Vlaamsí schihürs in dt omgeviiig Van de grafe meesters), Freren, 1988,
15 Vlieghe 1993 (as in note 10), p. 303.
16 Haus Vlieghe, Gaspar de Crayer, Sa vie a ses Oeuvres (Mono-graphics du Nationaal Centrum voor de Plastische Künsten van de
XVlde en de XVIIde Eeuw, IV), 1-11, Brussels, 1972 and Hans Vlieghe, 'Gaspar de Crayer: Addenda ct Corrigenda', Genlse Bijdragen tot de Kunstgischiedmis, XXV, 1979-1980, pp. 15S-209.
17 For Vinckenborch, see Hans Vlieghe, 'Rubens' begin-nende invloed: Arnour Vinckenborch en het probleem vail Jordaens' vroegste tekeningen', IXederlands Knnsthistvrischjaar-hoek, XXXVIII, 1987, pp. 383-396; for Wolffort, see Hans Vlieghe, 'Zwischen Van Veen und Rubens: Art as Wolffort (1581-1 6-11}', Wallraf-Richartz-Jahrhuch, XXXIX, 1977, pp. 93-136 and Hans Vlieghe, 'Nog wat over Wolffort, zijn Atelier, en Van Lint', in Feestbundel bij de opening van het Knlvenierihof'en Rubenianum, Antwerp, 1981, pp. 83—87.
18 Jonathan Richardson, Two Discourses. I. The Connoisseur: an Essay on the whole Art of Crilksim a it relates to Painting. 11. An Argummi in behalf of the Science of a Connoisseur: Wherein is shewn the Dignity, Certainty. Pleasure and Advantage of it. London, 1719; for an excellent discussion of Jonathan Richardson and the problem of connoisseurship, see now Carol Gibson-Wood,_/«Hrfffe« Richardson. Art Theorist of the English Enlightenment, London-New Haven, 2000, the culmination of many decades of work on the topic. See especially Carol Gibson-Wood, 'Jonathan Richardson and the Rationalisation of Connoisseurship', Art History VII, 1 984, pp. 38-56.
19 The essence of the Fitst Discourse is evident from its subtitle, viz. 'Shewing how to judge I. Of the Goodness of a Picture; 11. Of the Hand of the Master; and III. Wheterh 'tis an Original, or a Copy'. See also Richardson 1719 (as in note 18), pp. 179—ISO (after a long discussion of the various divisions of connoisseurship); all well discussed by Gibson-Wood 1 98-'i (as in note 18).
20 Cf. note 23 below; indeed rhe full title of the Second Discourse, 'A Discourse on rhe Dignity, Certainty, Pleasure and Advantage ot the Science of a Connoisseur', is significant enough, especially when read in conjunction with the passage in which he note s that 'if Gentlemen were Lovers of Painting, and Connoisseurs, it would be of great advantage to the Publick, in 1. The Reformation of our Manners. 2. The Improvement of our People. 3- The Increase of our Wealth, and with all these of our Honour and Power' (Richardson 1719 [as in note 18], p. 41).
21 One les.s-commenced area in which connoisseurship maybe of undisputed use and interest is its Utitlity in identifying partnerships between artisis — especially painters — within a single work, as, for example, in the case of the great painting of the Beast of the Gods by Bellini and Titian in Washington, ot the much lesser one of the Wolf Hunt by Rubens, Snyders and Wildens in New York. Of course if one could establish with certain:}' the division ot labor on the Ghent Aitarpiece - now that would be an achievement!
WHY CONNOISSEURSHIP MATTERS
22 Nelson fionilm«. I tmgmtgn if Art. Am Approach to a Tbmry 4SjmUi. Indianapolis. 1976.
23 Cf. lUdufdson 1^19 (as in note 18), II, pp. 3-4, 40-42; for the long history of this tradition, beginning, say, with Henry Peacham's arguments in The Complect Gentleman of 1622, for example, see now Gibson-Wood 1984 (as in none 18), pp. 41-44.
24 Giovanni Morelli, Italian Masten in German Galleries. A critical history of the Italian pictures in Munich, Dresden, Berlin, London, 1883 and Giovanni Morelli, Kunstkritische Studien über Italienisch Malerei. Die Galerien Borghese und Dorla Pampbi-Iff tn Rom von Ivan Lermolieff{pseud.), Leipzig, 1890. See also Jaynie Andersons 1991 edition of the first of these texts: Giovanni Morelli, Delia Pittära haliana. Studii {sic} storico-critiri. Le Gallerte Borghese e D'jria Pamphili a Roma; a ciira di Jaynie Anderson, Milan, 1991.
25 Carlo Giniiburg, 'Clues: Rook of an Evidential Paradigm' in Clues, Myths and the Historical Method, Baltimore-London, 1989, p. 97.
26 Carlo Ginzburg, 'Inrer/disciplinarky. Vetoes and Compatibilities', Art Balletin, LXXVII, 1995, p. 534.
27 Martin Schongauer, Studies of Peonies, ca. 1473. For the attribution and its significance, see Fritz Koreny 'A Coloured Flower Study by Martin Schongauer and the Development of the Depiction of Nature from Van det Weyden to Dürer', The Burlington Magazine, CXXXIII, 1991, pp. 588-597.
28 Ginzburg 1995 (as in note 26), p. 536.
29 As an interdisciplinary discipline that nevertheless retains i cs disciplinary auronomy.
30 One has only to think of the incorporation of techniques such as infra-red reflectography, autoradiography, spectrogra-phy, in addition to rhe now srandard use of e-ray photography, into an abundance of widely used recent publications, beginning, perhaps with the first three volumes of rhe Rembrandt Research Project (A Corpus of Rembrandt Paintings, I—III, ed. Joos Bruyn etai.,The Hague, 1982-1989), and extending into catalogues such as the and the series of catalogues and monographs written and edited by Maryan Ainsworth leaning heavily on her work with infra-red reflectography (eg. Maryan Ainsworth, Petrus Christus: Renaissance Master of Bruges, New York, 1994 and Maryan Ainsworth, Gerard David. Purity of Vision in an Age of Transition, New York, 1998, to mention only those on her work on Petrus Chrisrus and Gerard David), as well as her exceptionally interesting catalogue with Egbert Haverkamp-Begeman chiefly on the paintings by Rembrandt in the Metropolitan Museum in New York using autoradiography (Maryan Ainsworth, Art and Autoradiography: Insights into the Genesis of Paintings by Rembrandt, Van Dyck andVermeer, New York, 1982).
31 Cf. Kathleen Weil-Gams Brandt, A Marble in Manhar-tan. The Case for Michelangelo', The Burlington Magazine, CXXXVILT, 1996, pp. 644-659 and Kathleen Weil-Gams Brandt, 'More on Michelangelo and the Manhattan Marble', The Burlington Magazine, CXXXJ.X, 1997, pp. 398^404, as well as many contemporary newspaper reports.
32 Cf. Matthew D. Lieberman, 'Intuition: A Social Cognirive Ncurosciencc Approach', Psychological Bulletin, CXXVI, 2000, 1, p. 109 for a summary,
33 'Although intuitions may often lead to suboprimal decisions, ic is still possible that intuitions are sometimes as good or better than judgments derived from deliberation. This quality of intuitions is nor necessarily a default circumstance due to deliberative strategies falling shorr when overused, but rather may be the result of the structural properties of intuition once it is considered in its proper information processing context*. Lieberman 2000 (as in note 32), p. 109.
34 Malcolm Glad well, Blink. The Power of Thinking without Thinking, New York, 2005.
35 Ibidem, pp. 3—8.
36 Brrenson 1920 (as in note 6); Max J. Friedlander, On An and Connoisseurship, Berlin, 1942.
37 Arthur S. Reber, Implicit Learning and Tacit Knowledge, New York, 1993, p. 5. Reber conrinued One of the core assumptions of our work has been that implicit learning is a fundamental, 'root' process, one that lies at the very heart of the adaptive behavioral repertoire of every complex organism'. For the original definition of rhe rerm, see Arthur S, Reber, 'Implicit learning of artificial grammars'. Journal of Verbal Learning and Verbal Bebaviw, VI, 1967, pp. 855-863.
38 For a good summary of the cognitive and neuroscientific aspects of this issue, see Lieberman 2000 (as in nore 32).
39 Ibidem, p. 110.
40 Richard Wollheim, 'Giovanni Morelli and the Origins of Scientific Connoisseurship', in On Art and the Mind, London. 1973, pp. 177-201.
41 For an important discussion of the role of the coup d'oeil -and more specifically the premier coup d'oeil in late seventeenth and early eighteenth century art theory - see the relevant pages in Thomas Puttfarken's book on one of the fundamental figures in the history of connoisseurship, Roger de Piles (Thomas Puctfarken, Roger De Piles' Theory Of Art, New Haven-London, 1985, especially pp. 96-99); Gladwell 2005 (as in note 34), pp. 44-45.
42 Antoine Bediara, Hanna Damasio, Daniel Tratiel and Antonio R. Damasio, 'Deciding Advantageously Before Knowing the Advantageous Strategy', Science, 275, 1997, pp. 1293-1295, and Antonio R. Damasio, Descartes' Error, New York. 1994, p. 212.
43 Lieberman 2000 (as in note 32), pp. 113-114; esp. 113: 'In general the basal ganglia learn temporal patterns that are predictive of events of significance, tegardless of predictor-reward relationship' (chough this definition may not always seem to apply to connoisseurs motivated by rhe usual kinds of rewards discussed earlier in this essay!).
44 Diego Gambetta, 'Can we Trust Trust', in Diego Gam-betta, ed. Trust. Making and Breaking Cooperative Relations. Oxford, 1988, p. 235.
45 Steven Shapin, A Social History of Truth. Civility and Science in Seventeenth Century England. Chicago-London, 1 994, p. xxvi
46 Sec especially Mario Biagioli, Galileo Courtier. The Practice of 'Science in the Culture of Absolutism, Chicago-London. 1993.
47 In addition to Shapin 1994 (noot 45), see also his earlier and equally fundamental work with Simon Schaffer, in Steven Shapin and Simon Schaffer. Leviathan and the Air Pump. Hob-bes, Boyle, and the Experimental Life, Princeton, 1985.
PART I - A TRIBUTE TO HANS VLIEGHE
48 Gloria Origgi, 'What does ic mean to trust in epistemic authority", paper first presented at the Italian Academy for Advanced Studies in America, on March 9, 2005 (forthcoming). I am grateful to Dr. Origgi for allowing me to make use of her important paper - never intended for use in connection with the science of connoisseurship — in this essay. Also relevant to the following discussion is Gloria Origgi, 'Is Trust an Epistemological Notion', Episteme, I, 2004, pp. 61-71.
49 Origgi 2005 (as in note 48), p. 21,
50 Origgi 2005 (as in note 48), pp. 19—24, Origgi reverses the order of these maxims, but I believe that a belief in rele-
vance precedes the commitment to the view that what the speaker has to say is not intended to be false.
51 Dan Sperber and Deirdre Wilson, Relevance. Communication and Cognition, Oxford, 1986, and Dan Sperber and Deirdre Wilson, 'Truthfulness and Relevance', Mind, 111, 2002, pp. 583-632 in particular.
52 Sperber and Wilson 1986 (as in note 51).
53 Olli Lagerspetz, Trust. The Tacit Demand. Boston, 1998.
54 Goodman 1976 (as in note 22).
55 Gambetta 1988 (as in note 44), p. 235.
56 Origgi 2005 (as in note 48), p. 22.
57 Ibidem,