La Tour’s abbé Deschamps
Catalogues raisonnés are in effect narratives, telling a story of how a work of art was created, more or less convincingly. They may on occasion offer alternative endings, but for the most part the convention requires the narrative to proceed in a simple historical progression (at least for each work: I have written before about whether the collection of such narratives for each picture may, in the case of a portraitist, not be more conveniently ordered differently, since no one expects readers to read the whole book at once). The result then offers the solution to a puzzle which is often more compellingly told as the historical events were unearthed in the present day rather than as they unfolded at the time. So without more apology let me give you an alternative account of the catalogue entry you will find for the fine portrait by La Tour now in the Louvre (above; J.46.162 in the numbering system used in my catalogue) of the abbé Claude-Charles Deschamps (?1699–1779), bachelier de Sorbonne, prêtre, chanoine et regnaire de l’église cathédrale de Laon.
You’ll find virtually nothing about him in the La Tour literature published before 1922/23, when suddenly an unknown “pastel par Maurice Quentin de La Tour” (left; J.46.1622) appeared. The only prior mention is in La Tour’s 1768 will (published by Maurice Tourneux in 1904) of “mon cousin Deschamps, chanoine à Laon”. There is no mention of Deschamps, for example, in any contemporary or subsequent biography of La Tour. By 1923, La Tour had become one of the great names in the saleroom, the dramatic prices achieved at the Doucet sale in 1912 having rehabilitated him to the top table. Even immediately after the war his value was huge: Wildenstein sold the président de Rieux in 1919 for 1.2 million francs (perhaps £3 million in today’s money, although the purchaser became bankrupt before payment was made).
The sale itself (Paris, Drouot, Baudoin, 16 March 1923, Lot 86 bis) was a little unusual: it was a single lot with its own catalogue (copies are very scarce, and not available online at the time of writing, but I’ve obtained a photocopy), presented by Baudoin and Martini, at the end of another sale.[1] The vendor (of this lot alone) was disclosed as “Mme veuve R…, de Douai”, and the catalogue contained some interesting details. The work was presented as “Portrait de Monseigneur Claude-Charles DESCHAMPS, Chanoine de Laon, cousin germain de M. Q. de La Tour”: we discuss the wording further below, but it derives from a handwritten label on the back, which read “Mr Deschamps, chanoine de Laon, cousin germain de Mr Delatour, le Peintre” and which has subsequently been lost.[2] The catalogue then offered some biographical details, provided by Lucien Broche, conservateur des Archives du département de l’Aisne:
Messire Claude-Charles Deschamps testa le 20 août 1779 et mourut peu après dans sa maison claustrale de l’ancienne rue des Prêtres, à Laon rue Sainte-Geneviève, … Le mobilier du défunt fut vendu, du 27 au 31 janvier et du 1er au 7 février, par le Greffier en chef du Chapitre de la cathédrale.
The reproduction in the catalogue is unusual in revealing that this loose sheet had been mounted in the manner of a drawing rather than a pastel, in a style that looks as though it had recently been done:
The 1923 catalogue also mentions that the pastel had been shown in the Louvre from August to October the previous year, included in the exhibition of La Tour pastels from Saint-Quentin repatriated from Maubeuge where they had been taken by the Germans and awaiting the reopening of the musée Antoine-Lécuyer.[3] Uncatalogued, the exhibition included only two other La Tour pastels not from Saint-Quentin: two masterpieces belonging to the Galerie Cailleux, then also on the art market.
The pastel, estimated at 12,000 francs, sold for 13,500 to the dealer Jules Féral, and soon after was acquired by the legendary collector and philanthropist David David-Weill. (At his sale in 1959, it was bought for a mere £900 by Harry G. Sperling, president of the New York dealers F. Kleinberger & Co. Five years later it was bought by Dorothy Braude Edinburg, a collector of prints, drawings and ceramics. The daughter of Harry and Bessie Braude and wife of Joseph Edinburg, an executive at the hardware firm in Boston of which her father had been president, she donated more than 1500 works to The Art Institute of Chicago, including this, in 1998.)
By 1926 the palaeographer and archivist Charles Samaran (presumably following a request to research Deschamps, perhaps from David-Weill – that isn’t clear) made some enquiries of Lucien Broche (unpublished correspondence, bibliothèque de l’Institut de France), but Broche, after checking with Charles Sorin, the archivist at Laon, was unable to locate Deschamps’s will or the other Deschamps documents from the bailliage de Vermandois which had been lost during the German occupation. (We shall see below why this doesn’t greatly matter.)
Despite sending a dozen works from his own collection to the famous 1927 exhibition of Pastels français du XVIIe et XVIIIe siècles (of the organising committee of which he was chairman), David-Weill did not lend this pastel. Instead a second version (the one now in the Louvre) had mysteriously appeared. It belonged to the artist (and son of a piano maker), Bernard Wolff (1860–1949), whose sister later bequeathed it to the Louvre. According to information provided by Élie Fleury to the authors of the 1927 catalogue, this pastel had been found at a “château du Boulonnais” by “un commissaire-priseur de Douai”; the date on which Wolff had acquired it was not given, but the pastel seems never to have passed through public sale.
Paul Jamot published both versions in a review of the exhibition in the Bulletin de l’art ancien et moderne in 1927:
By then the David-Weill version had been reframed, more convincingly. Jamot struggled to explain the relationship between the two portraits: it was clear that they were the same image, and that the smaller sheet couldn’t really be regarded as a préparation for the Wolff pastel. Although La Tour often made different versions using varied techniques, these usually followed the trajectory from préparation to finished work. Jamot, comparing the two versions with the other La Tours in the exhibition, saw the Louvre version as the outlier, failing to note how much later it was in the artist’s œuvre. Modern photographs of the detail of the face make the differences even clearer:
The evident difference in handling Jamot suggested might be explained by wear in the David-Weill version: but this isn’t particularly evident (although of course extensive later restoration might be an explanation). Suzanne Folds McCullough, as recently as 2006, sought instead to resolve the puzzle by dating the Chicago sheet to 1779 – ignoring the fact that if anything, the abbé looks younger in that version, and the 1768 of the Louvre version is practically at the end of La Tour’s activity. The particular difficulty is that there is no coherent narrative in which La Tour would do two such different, finished portraits at the same time: and yet the fall of light, shadows etc. show these to be versions of a single image.
But the real point was that while the Louvre pastel is a work of extraordinary boldness and virtuosity, the David-Weill sheet is relatively lifeless and passive. In my view this is because it may be a copy; it attracted attention in 1923 because the original was unknown (and because La Tour is such a formidable artist that a good copy of his work, made by someone evidently familiar with his earlier technique, can fetch a powerful punch, even if it is knocked out by confrontation with the original), and, once accepted and acquired by an illustrious collector, no one had (or has had until now) the courage to draw the obvious conclusion.
Ultimately this is a question of connoisseurship.[4] Many will not accept my personal opinion, which dissents from the views of other writers (up to and including the 2018 Louvre pastel catalogue, where the Chicago sheet was reproduced as by La Tour without qualification). And I can see that there are arguments in its favour – the strongest being that it is not an accurate copy of the Louvre pastel. But I wonder whether the Chicago version would have had unanimous support had it emerged after the Louvre version was known. Readers of this blog are of course invited to offer their own views which I will receive with interest.
There are two objective facts that may help my view gain wider acceptance.
Firstly, no one seems to have noticed the extraordinary coincidence of the two versions both passing through Douai, when there is no obvious connection between the sitter and his family and that town during the sitter’s lifetime or immediately following period. Had the two versions always been together in the same family, I think we would have been told. But the Louvre sheet came from a chateau in Le Boulonnais, quite far away from Douai where the commissaire priseur who handled it lived, and where the Chicago sheet originated with the untraceable “Mme veuve R…, de Douai”.
The second point is that the inscription on the David-Weill version (which was said to be in a contemporary hand) identifies the sitter as La Tour’s “cousin germain” (first cousin). This has subsequently infiltrated the literature, and is repeated in all sources including Salmon’s 2018 Louvre catalogue. But it isn’t true. As I demonstrated in 2016, Deschamps was in fact La Tour’s second cousin: he was the son of Denis Deschamps, maître écrivain à Laon, and Anne-Françoise Caton. The connection to La Tour was through Caton’s mother, Marguerite Garbe, whose sister Marie married the pastellist’s grandfather Jean de La Tour in 1669 (you can enlarge images on this blog by clicking on them):
La Tour himself described him only as “mon cousin”, in the 1768 will, a term he uses in the broad sense. So the writer of that inscription made a guess that at the very least puts him at some remove from the immediate family. The suspicion arises (which, now it has vanished, may be unverifiable) that the inscription is a much later addition (reinforced by the phrase “Delatour, le Peintre”).
And so even the question of the date of the Chicago sheet seems to me open. It may be that scientific studies of the paper and materials might yield an answer, although (even if my suspicions are well founded) a copy of this quality is likely to have been made with carefully chosen media. (Pigment analysis in pastels is much less advanced than in oil painting, and most materials used in the early twentieth century were also in use in the eighteenth.)
Turning now to the beautiful work in the Louvre (at the top of this post), there is no question about its authenticity (pace Jamot). It has been the subject of conservation in 2004 and 2013 (by Valérie Luquet, Marianne Bervas and Sophie Chavanne) and a detailed technical report was compiled in 2014 by Pascal Labreuche, noting among many other things the relatively modest materials used, and even the small droplets of fixative which are still evident on the surface.
As I have discussed in previous blogs and in my catalogue entry, the discussion in the Louvre 2018 exhibition catalogue where it was most recently shown failed to take note of my 2016 genealogy (or of the other new points in this post). It also offered no provenance for the Louvre pastel before the acquisition by the commissaire priseur de Douai as reported by Fleury beyond the assertion that it had been the “propriété du modèle”. That was presumably inferred from the inscription on the back, to which we return below, but it may well be that the search for provenance (to resolve the claims of the two versions) was what motivated the 1926 enquiries by Charles Samaran mentioned above.
It is also worth noting that La Tour preserved his contacts with members of his extended family throughout his life. When the artist’s much-loved brother Charles died (3 July 1766), La Tour was out of the country (in Holland), and Deschamps signed the burial entry at Saint-Germain-l’Auxerrois. He was, as noted above, mentioned in the artist’s 1768 will, the same year in which the portrait was executed.
Also mentioned in La Tour’s 1768 will was Deschamps’s sister Marie-Jeanne, Mme Mauclerc, and (like the abbé’s will which Samaran sought without success), her inventaire is actually at the Minutier central in Paris.[5] Mme Mauclerc died in her brother’s house, rue des Prêtres (now Saint-Geneviève), Laon on 22 September 1774 (attended by the abbé, but not by her husband). I have transcribed the inventaire in my chronological table of documents, at 10 January 1775, and it reveals among other family portraits in oil, “un autre petit tableau de forme quarré peint en pastel sous verre le quel représente led. S. abbé deschamps” – surely the Louvre pastel which she had evidently been given during her brother’s lifetime:
Her effects were divided among her siblings. (Here is a link to the Deschamps genealogy.) Deschamps himself died in the same house in Laon, 18 December 1779. In his own will (which I did manage to locate, in the Minutier central in Paris, with certified copies of other documents from Laon), he left everything to his niece Charlotte, Mme Dorison (another La Tour sitter, J.46.1631).
Returning to the rather faded inscription on the back of the Louvre pastel, there are several more puzzles to be solved:
which we transcribe as:
Claudius Carolus Deschamps Presbyter/Sacræ facultatis parisiensis baccalarius theologus/ecclesiæ laudanensis canonicus <mot rayé ou illisible> regnarius/anno ætatis 69/1768/DD Quentin de La Tour, regius pictor academicus, fecit
Whose writing is it? What does it say (in particular what is his age)? What has been erased and why? These difficulties are compounded by the fact that writing with a quill pen on untreated wood with prominent ring patterns is quite tricky: ink runs, and the pen is redirected by the unevennesses. The writing is certainly not La Tour’s own hand. But we do have a number of samples of the abbé’s own writing, as he was curate at Saint-Médard, Agnicourt from 1729 to 1744, and the entries in the parish register are in his hand. I think the fit is good enough to say that the inscription is his with reasonable confidence.
What does it mean? DD is easy enough, Dono Dedit (gave as a present) etc. “Ecclesiae lauanensis canonicus”: chanoine de l’église de Laon. But “regnarius” is a sufficiently unusual term that Jean-François Méjanès transcribed it as “regularis”. In fact “regnarius”, or in French “regnaire” or “renaire” (you won’t find either word in the Dictionnaire de l’Académie francaise), is a specific dignitary at the chapter at Laon, a kind of master of ceremonies (before whom a sort of sceptre was carried in solemn processions), and we know from various documents that Deschamps was Regnarius at Laon by 1774.[6] What I have been unable to establish is precisely when he was elevated to this rank, or whether prior to this he was the “sous regnaire”. There is thus a simple explanation that the erasure is of the word “sous” and occurred on his promotion to the higher position.
Trickier is the question of the dates. 1768 is I think unambiguous, but I’m not so sure that the age preceding this is definitely 69. Nevertheless the entire literature has inferred that the abbé was born in 1699 (simply by subtracting 69 from 1768). The phrase “anno ætatis” is much abused in portraiture, by artists and art historians alike who forget it means “in the [69th] year of his age”, i.e. he was 68 at the time, and so logically (depending on what day in 1768 the portrait was made) could have been born anywhere between 2 January 1699 and 31 December 1700. But is that final digit 9 at all? As you can see from some examples in the abbé’s hand from the Agnicourt parish register, it’s not impossible that he wrote 5, consistent with a birth in 1703/04 (left to right: Louvre inscription; 1735; 29e).
The reason the question has attracted my attention is that his parents, Denis Deschamps and Anne-François Caton, were married in Laon on 19 February 1703:
There is no indication that his father had been married before 1703 (although registers are missing for Vailly-sur-Aisne where the Deschamps family originated), and numerous documents explicitly describe Claude-Charles as the “frère germain” of the daughters of Denis and Anne-François Caton (including Noëlle, Mme Augustin Masse and Marie-Jeanne, Mme Pierre-Marie Mauclerc, mother of Mme Dorison); he is also described as a cousin germain in the registre de tutelles for Henry-Pierre Messager, son of Anne-Françoise Caton’s sister. The division of property recited in the inventaire of his sister Mme Mauclerc (1775) makes it quite clear that the abbé was her full brother, while Mme Berthelot (a half-sister by a later mariage of Denis Deschamps) is distinguished as a “sœur consanguine”.
Thus, if the inscription on the Louvre pastel is correctly “ætatis 69”, the abbé was born illegitimately to Denis and Anne-Françoise before their marriage (when Denis remarried in 1739, another child was born less than two months later; but four years before marriage is improbable, particularly since Denis Deschamps and Anne-Françoise lived in different towns before their marriage). If incorrect, his birth was unrecorded (improbable: the record of Anne-Françoise’s annual births at Saint-Cyr, Laon is continuous to the end of 1705). (Incidentally the problem isn’t solved by assuming the abbé had forgotten his own age.)
But there is an alternative, if surprising, explanation: here is the entry for the first child, baptised Claude-Charlotte on 17 November 1703:
I have found no entry for the death of this girl or any other record of her existence. Is it perhaps possible that the future abbé Claude-Charles was misidentified as a girl at birth? The child was baptised the day of its birth, somewhat hastily (baptisms were most often the day after birth unless the infant looked as though they might not survive until then). Seven out of eight of his siblings were girls. Mistakes of gender at birth were not such an unusual occurrence; in 1731 one of the twin children of Jean-Antoine Philippe, another La Tour subject, was wrongly registered. That would be consistent with a reading of “ætatis 65”.
Whatever the abbé’s age, we can but agree with Jean-François Méjanès who commented of this, the latest of the Louvre’s La Tours, that the restrained palette of the pastel strokes “accentue néanmoins l’intensité expressive du visage sur lequel s’est concentré l’artiste”; the “grande attention” and “profonde humanité” that emerge justify more than any of the other works shown in the La Tour exhibition of 2004 the title of “voleur d’âmes”. Those who visited the Louvre’s pastel exhibition in 2018 will have been able to form their own view (although not perhaps as closely as they might like, as this small jewel was skyed in the hang I have discussed elsewhere).
NOTES
[1] That of Paul-Émile Rémy-Martin, the second of four sales of the collections that the cognac merchant’s father, Paul-Émile-Rémy Martin (yes the hyphens are in the correct places), had assembled at the château de Lignières.
[2] I am most grateful to the Art Institute of Chicago for confirming this to me, and the absence of any image of the lost label.
[3] See Fleury 1922b; Cabezas 2009a; Prolegomena, §xii.6.
[4] Those of you familiar with the system in the Dictionary will see (from the absence of the Greek letter σ) that I have not inspected the Chicago sheet in person. While I would prefer to do so, the factors I have taken into account in reaching my view are unlikely to be altered by examination de visu.
[5] AN mc lxv/386, 10.i.1775
[6] This is in the power of attorney he granted on 22 October 1774 attached to his sister’s posthumous inventory. In the Bulletin de la Société académique de Laon, 1913, André L’Éleu published a commentary on Fromage de Longueville’s unpublished Entretiens (1765), which contains encrypted satirical portraits of his contemporaries including one “Erophile, chanoine regnaire”, whom L’Éleu identified as Deschamps, but I suspect anachronistically. I have been unable to find detailed records for the chapitre de Laon before 1768.
Neil Jeffares 