SPECIAL EDITION

MICROSTORY OF ART ONLINE JOURNAL FOR ART, CONNOISSEURSHIP AND CULTURAL JOURNALISM A New Salvator Mundi Questionnaire

(10.3.2023) I am enlisting questions here that I am asking myself. Not for the first time, actually for the fourth time I am enlisting Salvator Mundi questions (and some of these questions have had very real consequences), but more and more disillusioned I am, as to a scientific community actually meant to tackle the Salvator Mundi problem as an intellectual problem. Is there anyone, in Leonardo studies, actually seriously working on this problem? I don’t think so. I don’t see anybody working seriously, and I don’t see any work which I consider worth to be taken seriously. Sad. But let’s transpose this disillusionment into a question: Is there anybody at all interested in the Salvator Mundi problem as an intellectual problem? Or is anybody only interested in treasure hunt, vague suggestions, and appearence instead of reality? In the latter case nobody is forced to continue reading. But anybody interested in the intellectual problem is invited to do so. The sad truth seems to be that the Salvator Mundi problem, having been produced by the academe (and not by the art market), is a problem that the academe either does not want to solve or is not capable to solve, and this means simply that the academe is not willing or not capable to solve a problem that it has created itself, and that the academe is apparently not able to handle questions of attribution satisfactorily. Who else is meant to handle such questions? After the academe, for many years, has suggested that it is meant to handle them?

1) Artificial Intelligence

Since a recent international conference had been meant to address Salvator Mundi questions, one might think that this conference would have been the place to raise questions as to the relevance of artificial intelligences in the field of attributional studies. But this was not the case. One of the big problems on the agenda has simply been ignored.
Journalists (with a half knowledge, at best) do seem to like the subject of artifical intelligences, apparently, solving or helping to solve attributional problems. Or at least suggesting to do so, or to be of help.
But some questions have to be asked, here and right now: If artificial intellegences have to learn first how to distinguish a Leonardo painting from anything else, is it possible at all to teach our robot friends? Or are the necessary preconditions simply lacking? Let’s say that Leonardo is an artist whose manner can be described in 100 aspects; and let’s say that his pupil and assistant and heir Francesco Melzi is an artist whose manner matches 80 of these aspects or is very similar. One could replace the name of Melzi with another name from the Leonardeschi pool, but I do think that Melzi is the most important and the most neglected name, both at the same time. How to teach our robot friend how to tell Leonardo from Melzi? Does a Melzi oeuvre catalogue with at least 200 paintings exist? No, it does not. No 200 paintings, no oeuvre catalogue at all. Does a Leonardo oeuvre catalogue exist? A Leonardo oeuvre catalogue on the same professional level as one existing Francesco Francia oeuvre catalogue does not exist. And scholars have not even begun to tackle these problems. Scholars have not even begun systematically to think about the problem of hybridity in paintings conventionally ascribed to Leonardo. On what basis, thus, our robot friend could be taught to tell Leonardo from Melzi? And I think that the answer is: presently it cannot be done. So, if we are told that artificial intelligences are somehow relevant to solve the Salvator Mundi problem presently, one simple question has to be raised: are these intelligences capable a) to explain to us in terms of language, based on what reasons the painting is classified in this or that manner? And b) do these intelligences pass the test at all to immediately indicate which parts of the Salvator Mundi painting have been repainted by the restorer, Dianne Modestini? Because we, we humans, do know, which parts, and this knowledge can be turned into a test.
The answer to the latter question is no: no artificial intelligence has yet been capable to immediately and precisely indicate the parts that have been painted anew by the restorer (the thumb of the blessing hand in the first place, but also the dark background). And if this would not be reason enough to immediately dismiss the relevance of such artificial intelligences, also the answer to the former question is no: no, these robot friends do not explain anything to us, so that we are not able to verify if a decision of classification has been done based on good reasons, or on reasons a human would be immediately able to identify as false and unacceptable reasons. Brief: artificial intelligences, in the context of scientific connoisseurship, cannot have a place at all, presently. Perhaps in the future, but presently – certainly not. And if anyone has solved the problem of how to tell Leonardo from Melzi with certainty and based on solid oeuvre catalogues, please do let me know. Certainty has been suggested by scholars and by the art market, but I do think both was just (auto-)suggestion from the beginning, and now too much is at stake to go back to the beginning, but exactly this has to be done, if we are willing to take the intellectual problem posed by the Salvator Mundi seriously.

2) Authenticity

It has become obvious to me in recent years that the first wave of Salvator Mundi scholarship is, I am choosing harsh words, not worth anything. The problem of telling Leonardo from Melzi (Luini, Cesare da Sesto, Giampietrino etc. etc.) has simply been ignored from the beginning. No effort has been made to address the problem (not to mention: to explain it to the world how exactly this problem had been solved in the individual case). And this is, concerning the intellectual quality, simply pathetic. Fallacies that are known since the 5th century BC have been ignored, and one such fallacy should be named: one does state that the painting in question matches some of the aforementioned 100 features that make Leonardo’s style, and for most people who want something to be true, this is enough. One does forget that many other artist also do match some features, many features, perhaps almost all features. And since the problem of how to tell Raphael from Antonio del Ceraiolo has been in discussion recently, we may observe that, again and again, it is the same problem: people seem not to wanting to think in alternative hypotheses; they forget to ask for the artist whose works also match many of the features that are characteristic for Raphael or Leonardo. And after scholarship has gone wrong for millions of times one should actually know about this fallacy, which has a name: confirmation bias, and characteristic effects (the neglecting of artists like Melzi or Antonio del Ceraiolo, and myriads of false attributions).
All arguments one had based the Salvator Mundi attribution upon have eroded in recent years (the pentimento argument is virtually falsified twice, and cannot be used anymore, if you are not able to refute the refutation; I am not saying it is falsified, but virtually falsified: eroded to the degree that it does not make sense to use it anymore, as, let’s say, ten years ago); the same applies to the argument that Hollar saw version Cook (the notoriously famous version) and thought it to be a Leonardo, and the same applies to the whole construct of a provenance narrative, which is simply obsolete if you are not able to refute my reasearch as to the Fontainebleau group.

3) Critical Thinking and the Establishment

Whoever scholar who denies reality disqualifies him- or herself as a scholar. If you are not willing to discuss refutations and to refute them, if you are not willing to discuss alternative hypotheses and alternative reasonings, you place yourself outside the academe. Scholarship is based on the distinction of true vs. false, but many scholars in the Leonardo field do not seem to know or to care about these basics. Journalists with a half-knowledge at best do seem to think that it is important what the Louvre says or what the Prado says, unaware that it is always about the reasoning and not about who or which institution does say something. It is understandable that journalists tend to turn to the people who actually should know the answers to all the questions raised here: but do the people meant to know indeed know? Does an academic publication which is marketed as the ›definite study‹ indeed deserve trust? Does it display critical thinking in the aforementioned sense? Or does it simply suggest that all problems raised here had been solved satisfactorily once, so that it is not necessary to rethink a narrative which, in fact, is eroded to the point of being virtually falsified, and which is not even represented in terms of one solid academic publication? If the academe is failing, there still would be journalism as a critical force. But journalism is largely busy to report on other glamorous or pseudo-glamorous treasure hunts, instead of tackling the intellectual problem.

MICROSTORY OF ART
ONLINE JOURNAL FOR ART, CONNOISSEURSHIP AND CULTURAL JOURNALISM
HOME

Top of the page

Microstory of Art Main Index

Dietrich Seybold Homepage

© DS