Super omnes docentes se intellexit…

Those are the words inscribed on the old frame of the famous La Tour Autoportrait au jabot now in Amiens (left). They are not directly from Psalm 119 (no. 118 in the Vulgate), but from St Augustine’s commentary, where the authorial voice (“me intellexi”) is turned into the third person. The King James version of verse 99 is: “I have more understanding than all my teachers: for thy testimonies are my meditation.”

I’ve been meditating quite a lot about this picture (and some testimonies about it) in the context of preparing my La Tour catalogue. The standard approach is to lock yourself away for many years and release the final product on paper. I’m trying to do this differently, sharing the work as I go, as you can find in the various documents on my website indexed from here. Sometimes I release a fairly final version (corrections are always welcome!) of my thinking, as in my recent entry on the portrait of Mme de Pompadour, which perhaps I should have shared on this blog. But the present portrait (or rather, group of versions of it) raises many issues which I haven’t fully resolved, and so the blog is the best possible way to share the puzzles and open the discussion before I go nap on the definitive cataloguing. You’ll see why if you read to the end. As much of the intermediate workings are rather detailed, skip straight to the end if you want.

Everyone will be aware that there are several versions of the Autoportrait au jabot (and at least ten later copies that make no further appearance in this post are listed in my online Dictionary of pastellists in the La Tour self-portraits article), and that one of them was supposed to be the one La Tour exhibited at the Salon in 1750 when he tricked his younger rival by placing it next to the pastel of La Tour himself that Perronneau had made – probably (or itself perhaps a version of) the pastel now in Saint-Quentin:

Too much has already been written on the respective merits of the two portraits. But while the mirror compositions suggest that the La Tour self-portrait, of all the known types, was surely that “au jabot” (rather than say the “oeil-de-bœuf” or “chapeau en clabaud” etc. types); that the specific work was that in Amiens (reproduced at the top of this post: no. J.46.1128) has rather been deduced from the fact that it is larger than the other versions assumed to be autograph, those in the musée Cognacq-Jay (J.46.113: left below) and in the Norton Simon Museum (J.46.1132: right below). Horridly I’m going to refer to these pastels as Amiens, CJ and NS.

This isn’t assisted by some erroneous conflations and confusions in the literature, so that, for example, Besnard & Wildenstein 1928 (p. 149) records Amiens as having a note on the back stating that it was made in 1751 – a year after the salon in which it was assumed to have been shown. (That label in fact belongs to a version sold in 1867 which it turns out is not the Amiens pastel at all; I list it as J.46.1131.) Among dozens of other errors in the literature I will mention here just two more: the date inscribed on the back of NS is 1754, not 1764 as usually reported; while CJ, contrary to Mme Burollet (Pastels et dessins, 2008, p. 139), was not the one from the Laperlier collection sold in 1879, lot 52 (that was NS) – CJ has no secure provenance before Pierre Decourcelle who sold it in 1911.

It was while I was trying to resolve the 1750/1751 confusion that I began to look harder at the questions these versions raise. I am most grateful to the curators at Amiens and Pasadena for providing imaging and documents that I discuss below. I should also remind readers about the usual important notice about attributions being subjective etc., and record the fact that the provisional suggestions I make below are not endorsed by other experts.

Before we get into the documents or delve further into the literature (even recent publications remain hopelessly confused), what can be said of the visual appearance of the principal versions? My own belief, before the recent discoveries, was that CJ was the best, showing all La Tour’s brilliance and inventiveness, while remaining an autograph replica of Amiens, which “must” be the one shown in 1750. I was a little surprised on the several occasions I saw it (in the musée de Picardie, Amiens and most recently in Orléans, when it was lent to the 2017 Perronneau exhibition) that Amiens seemed underwhelming for the mythology attached to the 1750 competition: as I wrote in a recent (but before the discovery at the end of this post) private email to a curator, “The Amiens pastel is not entirely happy: the jabot always struck me as a little pedestrian, while the shadows on the underside of the arms I find particularly perplexing.” (The shadows consist in some odd strokes of heavy black pastel.) But not to the point of questioning Amiens being autograph. Nor as far as I am aware has it been questioned by any other art historian (Fleury 1900a notes that he had not initially been persuaded of its authenticity, but changed his mind when the inscription and provenance were revealed), despite extended discussions in numerous sources (see the Dictionary entry for the full literature): thus for Debrie & Salmon 2000, it is “une œuvre essentielle”; while, in his préface to Dominique d’Arnoult’s Perronneau monograph of 2014, Xavier Salmon was even more emphatic, writing that the La Tour pastel exhibited in 1750 was “très certainement celui aujourd’hui conservé au musée de Picardie à Amiens, œuvre magistrale de psychologie et de maîtrise technique.”

CJ is smaller than Amiens in that the lower part of the bust is cut off; there is still space above the head. It is highly finished, with a superb sense of modelling which you can perhaps see most easily in the structure of the eye socket. The handling is relatively free – La Tour recreates effects rather than repeating each stroke exactly – just what I’d hope to find in an autograph replica.

NS (which I have not examined de visu, let alone side by side with the others – something which is not likely to be possible) caused me some concerns in the way it followed Amiens. While sticking to more or less the size of CJ, the figure is moved up so more of the bust shows, with less space above the head (see my scaled composite):

NS then imitates the exact composition of Amiens far more closely – for example, the angle of the arm, which in CJ is allowed to drop vertically, follows the angle of Amiens exactly: indeed the top of the hand placed in the waistcoat is still included, although it now makes little sense and might comfortably have been omitted had the artist allowed himself the same freedom as taken in CJ. There are differences too in the eyes: those in CJ engage us directly; those in Amiens and NS both seem slightly to veer off to the left. (You might think this an error in Amiens, corrected in CJ, so it is odd to find them repeated in NS.) More obviously the technique differs, in the face in particular, with a network of hatching in place of the finished appearance of CJ and much more prominent than in Amiens (although this isn’t immediately evident from the photography which makes the hatching on Amiens more prominent than I recall from direct examination). But elsewhere there is a very precise replication of each chalk stroke in Amiens: it is perhaps too close (in a way that is found in some otherwise excellent copies of other La Tour pastels).

I’m not worried about the appearance of these visible strokes on the flesh, per se; La Tour adopted this technique frequently, particularly in portraits intended for connoisseurs who he thought would be more receptive to the brilliance of these strokes which require to be viewed from a specific distance, while the general public found them too sophisticated. Perhaps the most extreme example of this heavy hatching is the pastel of Chardin in the Louvre (J.46.1436) from the 1761 salon. There are other examples from the mid-1750s – and, to make this problem even more tricky, there is very little sense of a chronological progression in La Tour’s technique that allows one to say that he used a specific technique at a particular time.

So it doesn’t follow that he came back to the Amiens pastel and made a replica say 14 years later in a different technique than the one he would have used say in 1750. You can’t even say that he wouldn’t make two versions in different techniques at the same time. But I can see why one would like to propose different dates for NS and Amiens and CJ, to help explain away the differences. Because of course the question is whether these are autograph versions. Normally the effect of these vigorous hatchings in the master’s hand is to make the portrait come to life. Judging from the photograph of NS I don’t have that immediate response. There seems to be a flatness to the modelling compared with the other versions that is surprising. Further the best of La Tour’s heavily hatched faces have an irregularity and energy I don’t see in NS. One shouldn’t attach too much importance to a single example, but putting the face in NS against a similar detail from his Chardin shows how differently he used this hatched technique:

Time now to broach the various inscriptions on the works. The reason we know it is NS rather than CJ that was in the Laperlier sale is because his 1879 sale catalogue mentions verses on the back by the abbé Violette, a priest in Saint-Quentin – just the sort of thing which makes you believe there is a continuous provenance back to the artist. So we have to look at that.

Although it wasn’t easy to find Violette’s biography, I can tell you (after a lengthy search of parish records) that he was abbé Charles-Théodore Violette (1737–1815), curé de Notre-Dame de Saint-Quentin, and a member of the Assemblée provinciale in 1787. So clearly it would be significant if NS turns out to have belonged to Violette.

As it happens there is rather a lot of writing on several different sheets pasted to the backing card of NS. There is a central panel in a mid-nineteenth century formal hand, with La Tour’s qualities and honours transcribed precisely from the title page of the abbé Duplaquet’s 1789 Éloge historique de La Tour:

Portrait
de M. Maurice-Quentin Delatour, Peintre
du Roi, Conseiller de l’Académie Royale de Peinture & de Sculpture
de Paris, et Honoraire de l’Académie des Sciences, Belles-Lettres,
et Arts d’Amiens, Fondateur de l’Ecole Royale gratuite de Dessin,
de la Ville de Saint-Quentin.
Peint par lui-même, en 1754

The “Peint par lui-même, en 1754” comes from this label alone. The writing is quite clear: 1754, not 1764.

The rest of the writing appears to come from a different, probably single, hand, but appears on two sheets. The lower sheet contains, on the left, an epitaph in Latin which I find extremely difficult to decipher. I suspect the Latin isn’t very good, but the text seems to echo the sentiments of the French verses below. Very broadly translated, it seems to mean:

This dust is mixed with the dust of Apelles, citizen of the Seine [“Sequana”] and citizen of the Somme [“Summa”]; they were astonished to find La Tour [“turreum”] under the soil; but his excellence and his mighty deeds will resound.

To the right are the abbé Violette’s verses which have been partially transcribed in several publications (but not the three lines at the bottom, which are critical):

Vers pour mettre au bas du portrait.

citoyen de la Somme, Apelle de la Seine,
de La Tour, dans ces traits, c’est bien toi ressemblant:
c’est ta bouche, tes yeux, ce rire caressant
qui vers toi tous les cœurs entraîne.
pour bien peindre le tien, ton âme, tes vertus,
bienfaisance, candeur, esprit, talens, droiture,
dons rares que te fit largement la Nature,
il faudrait toi, mais tu n’es plus!

L’inscription manuscrite, l’épitaphe et
et les vers français sont de M. Violette cure
de Notre-Dame de St quentin

As I read these, I don’t think the writing can be that of Violette himself. While sometimes people refer to themselves in the third person, that really doesn’t fit here, and he would probably have signed if it was his own writing. The verses themselves must have been composed after the artist’s death, but the inscription could have been transcribed by anyone who had found it at any time in the nineteenth century or later. It isn’t even sure that they were intended for this version – they could equally have been intended for the Amiens version, or indeed any other La Tour self-portrait (or any portrait of La Tour at all).

After La Tour’s death there were memorials (such as Duplaquet’s éloge, cited directly on the label), epitaphs and statues etc., so there were many occasions for the local curate to produce some verses of this kind. (For the very complicated events concerning La Tour’s death and burial at Saint-Quentin, see my La Tour documentation. The two witnesses were La Tour’s half-brother Jean-François de La Tour and the latter’s cousin, Adrien-Joseph-Constant Duliège, who as it happens was vicaire at Violette’s church of Notre-Dame. We met him in my last post, on La Tour’s brother and the letters that had descended to Mme Sarrazin.) But I don’t think that Violette was ever the owner of the pastel: the words could have been added later by anyone coming across his verses – perhaps even taken from another version.

But it turns out that there is more to learn from the upper panel, apparently in the same hand, which appears to contain some innocuous biographical information:

Maurice Quentin de la Tour,
Né à S^t Quentin, le 4 7^bre 1704,
revenu audit lieu le 26 Juin 1784,
ou il est mort et enterré au cimetière
de la Paroisse de S^t André, le 18
février 1788 –

Again these appear to be facts which would have been well known to anyone in Saint-Quentin throughout the nineteenth century or later. (The 26 Juin 1784 date is difficult to read; the month is correct, but the day should be 20 June according to other documents you can find in my chronological table of documents http://www.pastellists.com/Misc/LaTour_chronology.pdf.) The inscription gives La Tour’s date of birth as 4 September rather than 5: such confusions are common in a Catholic country where children were usually baptised the day after their birth, although in La Tour’s case he was born and baptised on the 5^th.

The significance is that the same mistake, and in fact exactly the same inscription, word for word (perhaps with misreadings: “revenue audit lieu le 21 juin 1784” and mort… “le 18 fev. 1783”), followed by “peint par lui-meme”, appear on the back of a miniature version of the autoportrait purporting to be by La Tour. At the time when it was described by Auguste Jal in his biographical dictionary, 1872 (sub verbo La Tour) it belonged to the princesse Mathilde (whom the Goncourt brothers derided for her susceptibility to fakes), and came from Aimable-Pierre-Joseph Opigez (1802–1881), a literary figure whose father and brother were alarmingly makers and retailers of objets d’art. It’s now lost (unless it corresponds with one in the musée Antoine-Lécuyer at Saint-Quentin: their two miniatures have no earlier provenance but I am investigating if we can tie them in). But as we know La Tour didn’t do miniatures.

The question however is whether this precise inscription taints the NS pastel or merely identifies it as the source of a later fake. It’s rather long to fit onto a miniature of normal dimensions. The possibilities include: (i) a common source for both NS and the miniature; (ii) the miniature copied from NS, or vice versa; (iii) Jal mistakenly referring to the princesse’s picture as a miniature when perhaps it was NS: but if so how did it get from her collection to Laperlier? (I couldn’t find either a pastel or miniature of La Tour in her posthumous sale in 1904.)

Having discarded the Violette provenance, the first certain sighting of NS is in the Laperlier sale of 1879 where the Violette verses are first mentioned. It is very probable that either NS or CJ is the pastel which belonged to Symphorien Boittelle (1813–1897), sous-préfet for Saint-Quentin before becoming préfet for the Aisne département, and later sénateur; in his sale at Paris, 24–25.iv.1866, Lot 70, not reproduced, is described as in a “light” blue coat, dimensions 44×35 cm. Boittelle’s collection was of mixed quality, and this was in one of a number of lifetime sales. The pastel reappeared with Jacques Reiset: his posthumous sale describes it as coming from the Boittelle collection, so it is no doubt the same. Both these sale prices were very modest, but that was a question of fashion rather than an indication of quality.

There was another sale in between, Paris, Drouot, Delbergue-Cormont, 8.xi.1867, where a pastel Lot 146 was sold, said to be dated on the back 1751 (although 1750 is mentioned in the preface), which all sources to date have identified as the Amiens version: the pastel is described but no size was given. It was said to be in a nice frame “en bois sculpté” (as CJ still is, while NS has been reframed). Although Amiens’s then frame was a fairly standard moulding which probably wouldn’t have been so described, and while its owners were attempting to sell it at that stage, a detailed analysis of the provenance shows that it cannot have been sold in the 1867 sale. (It is most probable that the vendor in 1867 was Sosthène-Louis-Félix Cambray (1819–1905), homme de lettres and a prolific collector and seller of drawings and prints. He might well have purchased Boittelle’s pastel. Although the commissaire-priseur’s copy of the 1867 catalogue shows Fr650 annotated against lot 146, it is not included in the list of bordereaux also bound into the same copy. However on the sheet opposite the lot is recorded “c.600 Lap.400 Gautier 300”, suggesting that Laperlier may well have bid, and perhaps bought it post sale, so this may well be NS – except for the 1751 date reported in the catalogue.)

In brief the 1867 pastel cannot be the pastel I’m calling Amiens because the musée de Picardie purchased that work (to which we now need to turn) in 1878 from the Lorne family who had owned it since 1796. It is true that the art critic Léon Lagrange had seen it around 1866 when the heirs of a previous generation were keen to sell, but it was not in fact sold then. I will spare you all the detailed steps from the 1770 gift of the pastel (recorded on another label pasted to the back of the work) by one Mlle Mangenot to the abbé Savary (he was Charles Savary (1731–1810), curé de Sainte-Colombe-lès-Sens) and its purchase in 1796 by François-Théodore-Clément Lorne (1768–1854), commerçant en gros de sel à Sens, who, the previous year, had married Savary’s niece but subsequently left it to his widow, his second wife: hours of harmless fun were required to establish these details, when of course, as Mme du Deffand would have told us, it is only the first step that counts: how it came into the hands of Marie-Louise Mangenot (1702–1782).

That takes us to her brother: the abbé Louis Mangenot (1694–1768), chanoine du Temple à Paris, poet, journalist, and great friend of the salon critic Philippe Bridard de La Garde who wrote so gushingly of La Tour’s later submissions. Mangenot was also the intimate friend of another La Tour subject, the playwright Crébillon (who made Mangenot his heir). Marie-Louise was probably the sister who was described in Palissot’s Nécrologe as“fort dévote”, but who “tyrannisait” her brother.

Most of the printed sources record Mlle Mangenot’s label, but only a few – notably Bitton 1936, which has been almost entirely overlooked – make the connection with her brother. Yet the abbé Mangenot does appear in the standard La Tour literature – as the author of yet more verses to be attached to a La Tour self-portrait:

Admirez jusqu’où l’art atteint
La Tour est gravé comme il peint…

Cited by Louis Hordret (Histoire des droits anciens… de la ville de Saint-Quentin, 1781), they were attached erroneously by B&W to the Autoportrait au chapeau en clabaud (J.46.1087). As that was only engraved by Schmidt in 1772 (after Mangenot’s death), it is however far more likely they were intended for the earlier Autoportrait à l’oeil de boeuf (J.46.1001), exhibited in 1737 and engraved by Schmidt in 1742. It of course, unlike the other self-portraits, does show the artist in his working clothes.

What this shows however is that Mangenot was indeed close to La Tour, followed his self-portraits etc. So there is nothing surprising in his owning one (except perhaps that there is no evidence that he had the means to pay for a major work by the artist). And one that he owned must be “right” in the sense that ones only traceable back to the mid-nineteenth century might not be – so that investigations such as opening the back and looking for anachronistic irregularities in the mounting of the pastel and canvas on the strainer etc. (often the easiest way to detect later fakes, of which there are sadly many in the La Tour catalogues) would be unnecessary (indeed pointless).

This is where things stood until a few days ago. While writing up my entries I investigated Mangenot more thoroughly, and came across this article in the Mercure de France, published in the edition for May 1755 (pp. 26-27). As far as I am aware it is completely unknown to art historians: indeed the only secondary reference I have been able to find (although not linked to Amiens, nor naming the copyist) is in the very useful Dictionnaire des journalistes in the entry on Mangenot (who would later edit the Mercure himself: it was then in hands of Louis de Boissy):

The footnote is, to say the least, astonishing. Unambiguously it identifies the La Tour portrait given to Mangenot as a copy by his pupil Jean-Gabriel Montjoie. Although he was mentioned in La Tour’s will and believed to be a pupil, recorded as an exhibitor in some minor events, virtually nothing was known about Montjoie’s biography until I unearthed some documents three years ago which are summarised in my Dictionary entry. Despite appearing in the Salon de la jeunesse in 1767 and later, he had in fact been born in 1725. But the surviving work, with one exception, all belonged to the 1780s or 90s, thirty years after Amiens. The one in the Louvre exhibited last year probably gives a fair account of his work: it was, I think we can agree, one of the weakest works in the show. The only earlier work I have found (J.543.11) is signed and dated 1768 – still some 15 years after the copy he made for Mangenot:

One isn’t likely to confuse this with the work of his master.

How do we make sense of all this? There are I think only two realistic logical possibilities, given that Amiens clearly belonged to Mangenot: either he subsequently acquired the original; or the Amiens pastel is indeed the copy Montjoie made for Mangenot as celebrated in the poem.

I’m not sure that either of these hypotheses will meet with universal approval. The first seems at best contrived, and raises all sorts of difficulties: why, having been given a version with which he was satisfied, would he seek to acquire the original – hardly likely that the artist would have made him a second present, so how could he afford it, and what did he do with the Montjoie copy? (The Lorne family papers make it quite clear that the abbé Savary had only this pastel, and 17 framed prints.)

The second requires us all to admit we were wrong in accepting Amiens as autograph. (The La Tour original, exhibited in 1750, must be lost – perhaps another work La Tour destroyed himself, or just still hidden away?) It requires us to reassess Montjoie’s competence. But is that such a step? Not only was Amiens made far earlier, when Montjoie was 30 years old and presumably at the peak of his skills; but it was also made under La Tour’s direction, and possibly with a good deal of assistance from the master. And the crucial fact we so easily forget is that it is far easier for artists to make brilliant copies of masterpieces than to create independent works of the same quality: we see this all the time in the pastiches and copies that flood the salerooms.

Do however look at the strange black shadows around the waistcoat buttonholes in the 1768 Montjoie. Isn’t that what troubled me about the arm in Amiens?

To sum up, provisionally at least. I’m quite happy that CJ is a fully autograph replica by La Tour. I have some doubts about NS which may nevertheless be fine: if at some stage it is opened for conservation I shall be interested in what can be seen, but I certainly wouldn’t reject it outright. Amiens in contrast, and to my surprise, seems to raise real difficulties: despite its quality, the conclusion from the Mercure footnote is hard to evade. Whatever assistance La Tour may have given, this appears to be the work of his pupil – you can only escape this by believing in a rather convoluted alternative narrative. The absence of the version shown in 1750, of which Amiens is no doubt a very precise copy, makes the classification of NS rather trickier as we cannot be sure how closely either followed the original.