The portrait painter is described in almost all art history sources as born c.1738 and still alive in Edinburgh in 1806 where he probably died. All sources correctly tell us that he was a pupil of Hayman, was in Rome 1758–59, and exhibited at the Society of Artists and the RA from Bath (1766), Edinburgh (1772) and India, where he was based 1776–85 until he returned to Edinburgh. There is a fine self-portrait (above) with a dealer whose website provides a fairly standard account of his career. The Oxford DNB has relatively recent entry which tells us that the portrait painter “(b. c. 1735, d. in or after 1806)”, “was born in Scotland, the son of Christopher Seton (d. 1768), a gem engraver.” As his paintings appear regularly at auction, the record ought to be corrected.
Seton was not a pastellist, so I know nothing about his work, but my attention was drawn to the 1815 will of one John Thomas Seton by art historian James Innes-Mulraine, who guessed this was the painter rather than a homonym (but why Kingston-upon-Thames, nine years after his last sighting in Edinburgh?), and wondered if it could be proved conclusively. It can, but not without a little work which I offer now.
The will itself is available on the National Archives website, or Ancestry, as a scan of the register copy of the proved will in the Prerogative Court of Canterbury. The will was made by “John Thomas Seton of Kingston upon Thames in the county of Surrey Gentleman” on 19 October 1812 and proved in London 20 January 1815.
There’s not much to go on – no mention of profession, indications of artists’ materials or other proof that this is the painter – and the handwriting is a bit tricky in places. What looks like a mention of Charles Grey of “Howick” turns out to be “Marwick” – nothing to do with the Earl of tea fame. The other beneficiaries, each like Charles Grey receiving £100 worth of stock, sound more promising: a Joseph Ward; Anthony Angelo Tremamondo (these first two named as executors); his son and Seton’s godson John Angelo Tremamondo; and one Mary Smith of Bideford; the residue to the family of his neighbours in Kingston, Philip and Susanna Pearce, whose daughter Elisabeth is to receive his “books paintings drawings and Instruments of Music”, as well as his lease of a tenanted property in Marylebone. There are mourning rings for “Lieutenant Colonel John Grey Eldest son of Charles Grey of Marwick” and to one Samuel Cotes.
There is nothing beyond the name to identify the last of these as the artist (brother of the better known Francis), but it is at least encouraging. As for the Smiths of Bideford, Mary’s father was probably James, a lawyer there, but his connection with the portraitist is unclear. Similarly the Pearces were real: “Susanna wife of Philip Pearce Gent” died in Kingston 31 January 1817 aged 67 – but again this gets us nowhere.
The Tremamondos – a fascinating family, far too interesting to attempt to sketch in a blogpost on another topic – had connections with Edinburgh and Bengal. The family name is Angelo, not Tremamondo, which makes locating Seton’s godson a little trickier. There is much material in several older sources, among them this and this. Anthony Angelo served in Bengal, was a friend of Zoffany and lived in Newman Street in the early 19th century: all good circumstantial evidence of a connection with our painter Seton. As for the “godson” John Angelo, he must be the John William Thomas born to Anthony Angelo and Martha Bland on 29 September 1792. Unhelpfully, and despite his names, he was baptised at St Patrick’s, Soho (3 October 1792) where the parish register lists as godparents Sir John William Rose, Dominic White Boyer and Rosalia Maggi – but not Seton.
So to complete the picture we must turn now to the usual online sources from which one can find so much genealogical material (but often not quite as easily as it looks). Wills in those days were usually proved very soon after the testator’s death, and there is no doubt that our will-maker was the “John Thos Seton Esqr” buried in the parish of Kingston upon Thames on 12 January 1815, “aged 79”:
That implies a birth in 1735, consistent with the artist’s presumed birth, which is highly encouraging (although as it turns out, not quite right). We know too that the painter was the son of the seal engraver Christopher Seaton (as he preferred to spell his name). A search of parish registers at the time yields only one possible baptismal entry, at St James’s Piccadilly, on 17 December 1733, the infant John Thomas Seaton being born on 1 December, to Christopher and Barbara:
Christopher and Barbara’s marriage is harder to trace, but the Faculty Office Marriage Licences for the dioceses of England & Wales list a licence granted 25 Feb 1731 to “Christopher Seton” and “Barbarah Davidson”. (Two years later he was living in Queen Street South, in the parish of St James’s, according to the Westminster Rates Books.) But still there is no proof we do not have a homonym.
The National Archives index one further instance for “John Thomas Seton”, namely a document in the London Metropolitan Archives evidencing that on 8 August 1792, “John Thomas Seton, 49 Upper Marybone, Marybone, gent” insured this property (it is near, but not the same, as the Marylebone property mentioned in the will). The address was given by some other artists in exhibition catalogues, among them Jane Read and Miss Betham in the early 19th century; by the engraver James Walker in 1783-84; and by a certain Thomas Mortimer in 1791.
All tantalasingly close, but not proof. To do that we need to go back to the Grey family – and by a lucky find, an account of a property transaction in two volumes of the History of Northumberland. In the 1893 volume 7, p. 410, from dealings in land in Felton parish we learn that John Thomas Seton of the will was in fact the cousin of Edward Grey of Alnwick (a relation of the Greys of Morwick); that he had lived in Calcutta but by 1801 resided in Marylebone:
The final piece of the jigsaw is in volume 2 of the same set, p. 460:
Here the pedigree of the Greys of Alnwick reveals that Edward Grey’s mother was Mary Davidson – evidently the sister of Barbara, Mrs Christopher Seaton, and the painter’s mother.
That’s as far as I need to go, although I am conscious of many loose ends (notably the art collector Alexander Davison) which I shall leave others to pursue.
Katie Scott & Hannah Williams, Artists’ things: rediscovering lost property from eighteenth-century France, Los Angeles, 2024. Available on paper from 9 January and online
A delightful – and quite unexpected, as I’d seen no publicity for it (although I knew that both authors have been working on it for some time) – discovery over the Christmas break has been the new book by Katie Scott and Hannah Williams which is published by the Getty. Paper copies of this nearly 400 page book will soon be available, but if you think the £50 price offputting, don’t be deterred – the authors, and the Getty, have generously made it available online, at no cost. You will all know that I am strongly in favour of disseminating knowledge this way: so many art books lie unread, or are consulted for a single fact before being put away for ever (or in my case buried under piles of other books never to be found again, so that even when I want to consult them again I have to find a copy elsewhere). Brave, if that is the right form, for this seasonal offering. (I’m putting this post up now so that you can enjoy it over the New Year break, even though the site may be an advance release – but there doesn’t seem to be any embargo.)
And it is something of a plum pudding: it had me sitting in my corner, putting in my thumb and pulling out endless fascinating stories about objects in the daily lives of painters – all connected with the French ancien régime. Not only is the book online, but you have a choice of modalities of access, including web-based browsing views by various taxonomies for the idle and curious. I confess (despite my protestations, I can’t help having been trained as a book editor half a century ago) that I rapidly switched to the pdf which I read sequentially (not always with even concentration), hungry to devour it all and anxious not to miss anything. I urge you to try both – and to stop reading this post before it gets too Beckmesserish.
So to the formula. If you read Williams’s brilliant article on “The mysterious suicide of François Lemoyne” (Oxford art journal, xxxviii/2, 2015, pp. 225–45) you will know already: take an object from the everyday life of an artist and probe deeply into every aspect of it to extract what one can about the owner. The aim is to avoid falling into the traps that so often catch conventional art history narratives, predetermined by generations of rigid thinking. I welcome anything that frees us from these straitjackets – such as this orthogonal approach that slices art history from a different angle. But ultimately the interest is only art historical – as opposed to social historical – if we can tie it back to the artist’s œuvre. Otherwise it could just be anyone’s inkwell or coffee grinder or kite.
The book does include Lemoyne’s sword (the object with which he killed himself), but adds another 54 or so objects. There is an intelligent discussion in the introduction on the approach taken and order chosen (alphabetical), acknowledging potential models such as Neil McGregor’s famous History of the World in 100 objects (2010) as well as the Encylopédie. At times I wondered if a closer model was not the Dictionnaire amoureux: Pierre Rosenberg’s volume dedicated to the treasures in the Louvre has a similar capacity to delight and enthrall. As a committed lexicographer myself, I’m all in favour of the alphabetical order for handling large amounts of information. I wondered here if the approach was altogether successful: there’s a good deal of repetition if you read it straight through, and only so much which the cross-referencing (the hyperlinks work well even within the pdf version) can do to alleviate this, when a single, thorough and logically structured discussion could be more efficient. But “logical structure” is precisely the railroad the authors seek to escape.
It’s curious too to see the discussion of the randomness of the order of items in Cesare Ripa’s Iconologie (a copy of which is one of the “things”), juxtaposing “Aurora” and “Avarice”, and one couldn’t help wondering if some of that criticism (pp. 50f) wasn’t autoreflexive.
The book is written clearly throughout and largely free of academic jargon (“The focus … is therefore phenomenological rather than psychoanalytical, trained not on interiority but on corporeal gesture and activity”, p. 7, is just about digestible). It achieves this without concession to academic rigour, and the book is safe in the hands of newcomers and specialists like, all of whom will find something new. Both authors write well, and while each notice is signed (discreetly), I found my success rate in guessing who had written which article was less than 100%.
What is also notable, if unsurprising, is that a good many of the “things” are objects that one would expect to find studied in conventional art history: books, diaries, documents, paintings, materials etc. The defence is that these are looked at from a different angle – thus Wille’s journal is discussed not from the point of view of the content (four of the five surviving volumes were published by Georges Duplessis in the nineteenth century and most of us have delved into his transcriptions) but from the materiality of the volumes as objects. So we are told that the five volumes were sourced from three different stationers (although we are not told that Jollivet, who supplied volume 3, was the brother-in-law of Bougy, who supplied volume 1). And while we are told about the reappearance of the missing volume in 2005, and given a reference to a “detailed analysis” of this, we are not told (unless I missed it) whether a transcription had yet been issued – probably the most urgent question that will come to readers’ minds if they have trawled through the other four volumes. As to the discussion about the gaps in the journals and how they fitted Wille’s biography, I found myself wondering whether this wasn’t part of conventional art history (how much of the analysis depends on the ink and spacing on the paper as opposed to the simple dates as printed in a transcription?).
I naturally turned first to the entry on “Pastels”. It proceeds from the virtual object in Mme Roslin’s lap in her husband’s oil portrait of her using them to portray Henrik Wilhelm Peill (the Nationalmuseum painting is so relevant you will even find it in my Prolegomena, §vi.1, fig. 2: the Prolegomena is not cited although it has a great deal that is relevant). The discussion then moves onto Vernezobre’s pastel box – which would have been a more obvious object to choose, except that the authors seem to have preferred to link each object (mostly) to a member of the Académie royale. Curiously the Académie de Saint-Luc, of which Vernezobre was a member, is never mentioned by name, but referred to consistently throughout all the essays as “the Guild”: if this was an effort to avoid confusion, I think it likely to have the opposite effect.
We swiftly move on to the posthumous inventory of Vernezobre’s wife, the reference to which given in footnote 11 is simply to the cote in the Minutier central; there is no mention of the fact that I published this document, with a transcription of all the art-history parts, in 2018: you can find the discussion (including the lists of those Vernezobre supplied, annotated with suggested identifications) on this blog. Were the authors really unaware of this, or did they feel it wasn’t worth citing? If so I hope that wasn’t because my publication was online: a view which I hope publications like this will help dispel.
Closely related to “pastels” is the discussion of “crayon”. Again this immensely intricate (and for most, tedious) discussion of linguistic usage is dealt with at length in my Prolegomena. Most of the discussion here is accurate, and the actual entry on Crayon focuses on the manufactured crayons marketed by Nadaux. Elsewhere however the usage seems to be inconsistent, and again I couldn’t help but feel that for readers who aren’t completely secure in understanding the difference between graphite, naturally occurring black chalk, artificial reconstituted crayons and pastel, a single discussion would be easier to follow. Incidentally, while I am cited for Nadaux, no biographical information is given for his inventor, “Gabriel Dumarest” [sic]. He too gets his entry in my Dictionary, not as a pastellist (there is no evidence he worked in the medium), but in my index of inventors, writers and suppliers http://www.pastellists.com/Suppliers.html#D, under Dumarets, although his real name was in fact François-Gabriel Pigeon, as he was baptised in Paris, 17 March 1721 (as far as I know that has not been published by anyone else). Once again the question that comes to mind in reviewing Scott & Williams is just how their approach illuminates art history better than conventional narratives: isn’t the idea to grab hold of a theme and push it as far as one can go, including unearthing new facts about the people involved?
Back to the title, or rather subtitle: Rediscovering lost property from eighteenth-century France. I was expecting to find lots of accounts of everyday objects lost in the street and recovered (or not) through advertisements. Indeed I devoted a post on this very blog to this idea eight years ago: Lost & found: some absent-minded pastellists, in which we rediscovered not only snuff-boxes and the like, but even a pastellist who had disappeared from art history. I certainly expected La Tour’s lost snuff-box, the subject of several legal documents (which of course you can find in the documentation section of my La Tour monograph), to make an appearance in Scott & Williams, but it doesn’t. (Of course my habit of reviewing publications not for what they contain but what I would have put in, unfair at the best of times, is especially unreasonable here.)
Instead we get the incident of Perronneau’s étui, which was first published in the nineteenth century by Maurice Tourneux and is in all the standard Perronneau sources. I published it with a corrected transcription of his advertisement, and you can see a facsimile of the entry in my 2015 blog post or here:
One of the (for me at least) annoying things in Scott & Williams is that translations rather than original texts are given, and here the source is cited as “the French newspaper Annonces, affiches et avis divers”: it is of course from the Annonces, affiches et avis divers, de la ville de Bordeaux – a completely different publication, if like me you enjoy seeing the original. Nor is the secondary literature cited of much help: neither Tourneux, Vaillat & Ratouis de Limay, d’Arnoult nor me, but an unpublished Ph.D. thesis by one of Scott’s former students.
In any case, what’s the point of the story? If it is to illustrate Perronneau’s absent-mindedness, perhaps the loss of the pastel of his wife (which I also discuss) would have been a better example. But the moral the authors extract, or at least the punchline (many of the stories end with such an envoi), is based on the idea that the étui contained a porte-crayon (yes), that the porte-crayon would have been used to hold the graphite with which Perronneau signed his pictures (debatable), and that
The artist Perronneau minus porte-crayon was, in his mind, and perhaps in the opinion of others, not just an artist with an incomplete tool set but, simply and more significantly, not quite an artist at all.
That to me is a step too far.
Some of you will have been perplexed by the caption on one of the pictures in that essay: a portrait of “Louis Metayer Phzn”; the suffix is just a contraction of the patronymic Phillipzoon, meaning son of Philippe [Metayer], but is odd outside a Dutch language context (such as the Rijkstudio website). Other name issues in the book I will pass over, except that we really should by now know that Mme Vigée Le Brun’s preferred forename was Louise, not Élisabeth.
Names do matter though: one of the real delights in the book was the discovery of the gaming set that belonged to Liotard. I was completely unaware of it, and as far as I can recall it is not mentioned in Roethlisberger & Loche or any other standard publication. Well done. Naturally my first instinct was to check whether it really had come from the family. We are told (twice) that this was presented to the musée d’Art et d’Histoire in Geneva in 1975 by “Marie-Margereta Liotard-Hülsche r” [sic]. Was she a direct descendant? You can spend some time on genealogy sites trying to verify this before correcting the three missprints and realising that she was Marie Margareta Hilscher (1891–1986); she married the artist’s great-great-grandson (Mme Tilanus’s nephew) Charles-Aimé Liotard (1871–1936). So yes the credentials check out. The essay analyses the gaming set in fascinating detail. As the game of Boston russe (one of the many whist-based games, a foreunner of bridge – but not one that makes it into Marcus Du Sautoy’s Around the world in 80 games) was not invented until the second half of the 1770s, however, Liotard would certainly have been very old when he received it. The book speculates as to whether Liotard would have used it – but surprisingly fails to mention his pastel in the same museum, his famous Nature morte : jeu de Lotto (J.49.2605 in my Dictionary) made before 1775 and including a much more ordinary bag of counters. One feels here that the omission, trivial in itself, was another lost opportunity to anchor the project to what really matters: artists’ pictures.
One final illustration of this dislocation and the unfortunate results of departing from conventional organisation. There are numerous mentions of the ordre de Saint-Michel, the highest chivalric order available to artists. It’s discussed under “Decoration”, although I’d have placed it under O or C (perhaps even B for badge or I for Insignia; such is the arbitrariness of labels). But it also crops up in at least eight other places in the book. And while the decline in the social weight attached to the order is discussed, nowhere (unless I missed it) is explained the concept of “les Ordres du roi” and how it technically fitted in with the Saint-Esprit – let me not take you through all that. But what is particularly surprising is that the portrait chosen to illustrate the main discussion, fig. 36 on p. 108, the Duplessis painting of Vien, accompanies a text telling us correctly that the insignia included “a black riband to be worn as a sash across the body and, hanging from it, a gold badge with the insignia of the order.” Yet nowhere (as far as I could see) is it pointed out that the Duplessis portrait shows Vien conspicuously not wearing the order, but having taken it off and hung it on the back of his chair – an observation the text repeatedly cries out to be made. (You can see the order correctly worn by Pigalle in Mme Roslin’s pastel reproduced in fig. 130 on p. 263.) I worry here that the visual evidence has been lost sight of.
All minor quibbles which shouldn’t get in the way of your enjoyment of a book into which a great deal of work has gone. It’s a delight from start to finish (if you read sequentially) or however often you dip in. And it’s an excellent vehicle to persuade students and public of the rich rewards such detailed multidisciplinary research can yield. There is room for many different approaches to art history, even if I’m not yet persuaded that this angle will become standard (perhaps the point is that it should not: it is its quirkiness that is its justification).
Apollo has chosen the NPG/Getty joint purchase of Reynolds’s portrait of Mai as its Acquisition of the Year 2023; I doubt if they had any difficulty making the choice, and there is no denying the effort and ingenuity that the NPG applied to bringing off this coup. You will recall that before the deal was finalised I wrote here about some of the difficulties to be confronted in such an arrangement. I have subsequently seen the joint venture agreement, which I don’t propose to analyse here in any detail, but suffice it to say that it doesn’t bring a magic wand to dispel the difficulties that may arise if the picture becomes unable to travel regularly, is destroyed in one or other location or in transit, or the parties fall out for any other reason.
I do however want to talk about the transparency of the process, and in particular the way the price was arrived at – the critical document being the independent valuation. Since a good deal of the £50m price was made up of public money or contributions from individual donors it does seem to me that the valuation procedure should be open to discussion rather than beng conducted sub rosa. We all know that it’s a good example by an important British artist – but that there is an added interest because it involves a person of colour, a theme which is currently in vogue. But how do those components break down? And there is a particular problem, acknowledged even by the chairman of the Reviewing Committee for the Export of Works of Art whose advice to the Secretary of State effectively set the price: in this case (unlike most the RCEWA consider): the picture had not been to auction, and there was no independent market confirmation of the price the owner sought. For that reason an independent valuation was sought, from Anthony Mould.
You may have seen an article in the Telegraph on 22 July (here, but behind paywall; you can read a version with ads here) where I am quoted; I’m not going to repeat or expand those arguments. But I do want to share the difficulties I encountered in pursuing this with the Freedom of Information Act.
First, and perhaps most surprisingly: the NPG do not have a copy of the Mould valuation, and apparently did not commission one of their own. They simply accepted that £50m was the price, and no further independent confirmation was required. (We should perhaps also note that (a point I made in my original blog), in a 50:50 joint venture, with the painting committed to travelling regularly, you are paying 50% of £50m plus an unending stream of transport and insurance costs for less than 50% of the painting’s wall presence (time in transit, conservation etc beyond what a straight purchase would achieve). So the deal effectively is at a premium to the headline price.)
Access to the Mould report then depended on the Arts Council’s response to my FOIA request, served on 2 April. It probably won’t surprise you that the process only concluded this week, with the final delivery of some minor documents as ordered by the Information Commissioner’s decision, issued 7 November. You can read the full decision here: I’m not going to summarise it, nor indeed to expand this blog with the many arguments and submissions I made regarding each of ACE’s claimed exemptions, which the ICO have simply omitted or sidestepped in their decision. This summary appears on the ICO website:
The complainant has requested information about a painting; the Portrait of Mai (Omai). Arts Council England (ACE) disclosed some relevant information and initially relied on section 40 (personal data), section 41 (information provided in confidence) and section 43 (commercial interests) of FOIA to withhold a valuation report and other information. ACE later also applied section 36 (prejudice to the effective conduct of public affairs) of FOIA to the majority of the information it confirmed that it’s withholding. ACE then confirmed that it’s withholding the remainder of the information in scope under section 21 (already accessible to the applicant), section 22 (intended for future publication) and 40. The Commissioner’s decision is that ACE correctly applied sections 36(2)(b)(i) and 36(2)(c) of FOIA to information it’s withholding under those exemptions ie: the Mould valuation report The majority of the information in Annex A1 (‘Annex A Information required by the Reviewing Committee on the Export of Works of Art and Objects of Cultural Interest in order to consider case referred’) and Annex B together with the Applicant’s application, Annex A2 (‘Annex A Reviewing Committee on the Export of Works of Art and Objects of Cultural Interest (RCEWA)’); and the Applicant’s valuation report. The majority of the information in the expert adviser’s statement, which forms part of Annex A2, is exempt under section 21(1) of FOIA; however, some of that information doesn’t engage section 21(1).The remaining information in the expert adviser’s statement isn’t exempt from disclosure under section 22(1) or 40(2). There was no breach of section 10(3) or 17(3) in respect of the timeliness of ACE’s response.The Commissioner requires ACE to take the following steps to ensure compliance with the legislation:Disclose the application questions in Annex A1 and Annex B.Disclose the small amount of information in the expert adviser’s statement in Annex A2 that the Commissioner has found section 21 can’t be applied to. This information is given in a Confidential Annex to this notice.Disclose the expert adviser’s name, role and institution in the expert adviser’s statement in Annex A2.
FOI 10: Complaint not upheld FOI 21: Complaint partly upheld FOI 40: Complaint upheld FOI 17: Complaint not upheld FOI 36: Complaint not upheld FOI 22: Complaint upheld
The outcome is that Mould’s valuation has not been released. I’ll come back to this in a moment. But it is worth noting that, for all its resemblance to a tribunal, the ICO does not operate in accordance with some of the rules of natural justice. The complainant doesn’t know the contents of the document he seeks, so can’t formulate his argument concisely and with focus; the ICO engages in private conversations with the public authority – to which the complainant can have no effective right of reply. (To rectify this one would want to see the case officer taking more of a role as advocate for the complainant, with the final decision being made by the Commissioner, or at least a different member of staff.)
In this case, exemptions under ss. 41 and 43 were claimed initially which didn’t seem to me to hold water; and I got the distinct impression that the original case officer told ACE they might not succeed, which is why they introduced a claim for a different exemption. Such a claim, under section 36 of the Act, needs (according to ICO guidance which the decision ignores, citing as precedent for doing so another decision made by the same case officer rather than a tribunal ruling) to be based on a “qualifying person’s opinion” issued before the internal review process is complete; in this case it was put together a month after the matter had been with the ICO. The decision (arrived at by a new case officer, but not appointed for the reasons I’m suggesting) simply failed to rule on the exemption claims originally made under ss. 41 and 43, and while they decided that the time taken to complete the public interest test was not a breach of ss. 10(3) amd 17(3), they were silent on the failure to comply with the basic requirement to claim a s.36 exemption within 20 days under ss. 10(1) etc. One could also challenge the rationality of the QP opinion which, like Alice’s White Queen, seems to have held two incompatible views at the same time (para 83: ICO guidance makes it clear why one must be selected, though this decision glides over the problem).
But ignoring all such legalistic points, what it came down to was a decision about whether disclosure of the Mould valuation was in the public interest. In my view the factors favouring disclosure present a clear case: a very large amount of public money was spent to acquire this picture on the strength of this valuation which has arrived at a number for which no rational justification has yet been published, and which is far beyond the level that any other picture by this artist has ever achieved. How this happened, what it means for UK museums seeking to acquire similar works and what it says about the mechanisms for saving works of art for the nation are all legitimate matters for public debate which cannot take place until the document is made available. Public institutions such as ACE, however reluctant they are to accept this, must be open to public scrutiny precisely as the Freedom of Information Act legislates.
It is also worth saying what I think the valuation ought to look like. It should present a reasoned argument, pointing to comparables in the public arena, and explaining how the various factors driving value in this case are to be combined. That way for example we can begin to understand what an equally outstanding Reynolds of a white sitter might be worth (a factor which might mitigate the inflation risk I’ve highlighted). It should also inform the understanding of the interaction of aesthetic merit and historical significance, a consideration of some relevance to future NPG acquisitions where mediocre portraits of significant people can have a place that would be denied them in say the National Gallery. It also should tackle the question of how any Reynolds could be worth a stratospheric price hitherto reserved for masters such as Raphael, Rubens or Rembrandt.
And also what the valuation should not look like. It should not merely be a hand-waving, stream of consciousness list of considerations in a half-baked contribution to debate – the kind of “free and frank advice” disclosure of which might have the “chilling effect” on discussion which s.36 is designed to protect. The public money spent on the (undisclosed) fee deserves something more finished and robust, something a competent valuer should be happy to defend in public.
Further in para 86 the ICO wave aside the point that the picture hadn’t been at auction, and conclude that the independent valuation was a complete antidote for this. That would only be the case if the valuation was robust, logical and unassailable. Remember too that the instructions to the valuer were to decide “the price which it might reasonably be expected to fetch if sold in the open market by a willing seller to a willing buyer who are both knowledgeable, informed, and prudent, and who are acting independently of each other.”
I see however from the snippets that appear in the decision (which don’t much resemble the arguments ACE submitted to me) that there are indications the valuation included some rather broad-brush handwaving. The ICO itself seems to have set aside its role in deciding whether due process was followed into itself assessing the arguments about value – a role for which it has no qualifications. It seems to accept the waffle ACE submitted in paras 92 and 93, that the premium is justified by the painting “being a portrait of a person who is other than white, at a time of British imperial expansion; a ‘Western’ style painting of a ‘non-Western’ sitter; and the intersection of science, technology, and progress.” (But how much? And wasn’t it applied the last time it sold, in 2001? And if not, isn’t this a temporary fashion which should be ignored by a “prudent buyer”?) Again we are told the five-fold increase since 2001 is explained by “the rise of terror related attacks in the UK and the world generally; the Iraq War 2003-2011; the Great Recession 2007-2009; the UK’s departure from the European Union 2016-ongoing; and the COVID-19 pandemic.”
All I can say is that if this vague drivel is the substance of the valuation, I am unsurprised that ACE fought so hard to ensure it is not published. It may of course have been an exemplary document meeting all my demands: I simply don’t know.
A number of conflicts of interest regarding the National Gallery’s new exhibition have made me decline an invitation to review it in a well-known art history journal. I’ve seen the exhibition catalogue, written principally by Fran Whitlum-Cooper (who has graciously sent me an advance copy; it may not yet be in general circulation). But there are a few points I’d like to share (more notes of what the exhibition doesn’t attempt than criticisms of what it does) which go beyond the very detailed analysis of the picture which I published some years ago on this blog and later expanded into an essay on my website (which you can find here), as well as a similar essay (with further material on the Lavergne family) on the pendant (here). Consult them for a full discussion with references etc. (Neither essay is included in the bibliography of the exhibition catalogue; one of them is cited in an endnote, but only as a genealogical reference which few will pursue. Among other omissions from the catalogue is a full account of the provenance which you will find in my essay.)
The most striking feature of the exhibition is of course the confrontation between the pastel and the oil versions (above is a photo, the colour inaccurate, that appears in today’s The Times; I return to the technical questions below). I suspect the overwhelming majority of visitors will have emerged with the simple verdict: the oil is better. (It was placed on the left of the earlier version, which seemed to me counterintuitive.) They will, I think, have had all their prejudices about pastel reinforced: that it is an anaemic medium capable of producing only feeble colours; the phrase “pale imitation” may even have been uttered by some. (I wrote about this linguistic issue ten years ago in a post on this blog entitled Polemic and the toxicity of pastel.) What many visitors may not understand is that most eighteenth century pastels have not deteriorated and many retain vibrant colouring, with higher pigment density than oil can achieve (there is nothing effete about the colouring of La Tour’s président de Rieux or Valade’s Loriot, figs. 20 and 21 in the catalogue); and that Liotard, through his self-taught technique, imperfect materials and taste for gingerbread colours (as Mariette called them), is simply not representative. An exhibition that confronted autograph versions in both pastel and oil including other eighteenth century artists from Perronneau to Vigée Le Brun would have left a very different impression.
There are of course many other fantasy exhibitions which the NG might have mounted around Le Déjeuner Lavergne. Perhaps most obviously by exploring the family whose name is retained in the title of the picture; but the catalogue includes very little information on them. My research has discovered several new siblings, amended their dates and suggested identifications of the sitters which are not to be found in Roethlisberger & Loche’s definitive catalogue. It is not merely that the girl is surely Anne Delessert, but that one of the candidates for the older woman is her favourite aunt Marguerite who left Lyon to live with Anne in Cossonay, where she died, aged 68, in 1795. We may never be certain of these identities, any more than that of the brother in the pendant L’Écriture, who may or may not be the one who threw himself out of a window in despair – a fact which led one critic to describe that picture as “tout imprégné de rêverie mélancolique”. By all means flag the uncertainty, but aren’t these the stories that can draw viewers into the pictures?
What we do know for sure about the Lavergne family is that Liotard’s sister married into a Huguenot network with connections across Europe, from Lyon to Geneva and Amsterdam, the very network that offered Liotard so much of his business. The Lavergne family (the catalogue simply identifies Sara’s husband as “a businessman”) were négociants en soie, later described as banquiers and commissionnaires – effectively agents who moved money and arranged trade for their clients – among whom they numbered Voltaire, who, his correspondence reveals, was well aware that one of the brothers was a gifted amateur playwright. One might also have explored the rather smart silk dress that the aunt wears – after all this was a family of silk merchants, in a town dominated by the trade in silk fabric (many pastellists, including La Tour, were brought up in families connected with the textile business). Perhaps some costume expert could take up the challenge of finding a similarly striped sample.
The important part of the genealogy I have set out in my essay is not the exact number of siblings or their ages, but the fact that Liotard painted so many of them (not just the obvious ones, but recent identifications include the brother-in-law of the aunt in Le Déjeuner Lavergne, Daniel Clarenc de Saint-Loup J.49.1262, whose name defeated R&L; and his granddaughter, Anne-Philippine Clarenc J.49.12631 recently bought by the Louvre, omitted from R&L; both below). The relations between the Lavergne, Delessert and Clarenc families matter to anyone studying Liotard.
Genealogy and chronology combine to produce art historical facts. Thus Marianne, depicted much earlier (1746) as La Liseuse (J.49.1765, Rijksmuseum), fig. 46 in the new exhibition catalogue, is not the elder sister Anne born in 1717 as the catalogue repeats (R&L were surprised she could be 29 but has no other explanation), but a younger sister actually called Marie-Anne whose baptismal record I found in the parish register: she was 12, not 29.
One of the unanswered questions in the catalogue is why Liotard interrupted his successful stay in London to travel back to Lyon. As I argue in my essay, that may have been because of the need for a family conference after the recent deaths of François Lavergne and of François Delesert, including perhaps decisions to be made on the little girl’s guardianship as well as on Étienne Delessert’s promotion to head of the business.
The still life, to which so much attention is devoted in this exhibition, includes not only the coffee set but a prominent sheet of music: this I suggest is an allusion to the fact that one of the Lavergne brothers was a leading light in the Académie des beaux-arts which held weekly musical concerts in Lyon. The other inspector was Claret de La Tourette (another Liotard sitter from the 1754 visit, J.49.1264).
The cover of the catalogue conforms to a series format: neither typography nor colour seem attractive. There is a logic to reproducing pastels on matt paper, but it rarely seems satisfactory in practice (I’ve written about this fashion before: see my Burlington Magazine article, August 2022, p. 783).
If there is no discussion of the Lavergne family, that was not for lack of space. The author, reviewing the 2015 Royal Academy Liotard exhibition in Journal18, commented that “it seems now mandatory to include a selection of pastel crayons and papers at every pastel exhibition: while this undoubtedly helps bring the medium to life, it does feel a bit incongruous to display a box of John Russell’s pastels at a Liotard show. One can’t help but feel that master of self-promotion Liotard would have been a bit miffed.” But this exhibition includes a tray of pastels by Henri Roché dated c.1910, and has an extended chapter running from Rosalba Carriera’s Uffizi self-portrait (which incidentally is not on paper) to Degas and Redon (all the illustrations chosen are well known).
While one La Tour is reproduced in the catalogue, there is no discussion of why Liotard was despised by the French art establishment (or why for that matter the English love him, then and now: why do we embrace Liotard, Pillement and Boilly but show little interest in La Tour and Perronneau?). That could have constituted another fantasy exhibition. The 2015 RA monographic show offered a broad overview of the Liotard’s career, within the limitations imposed by the medium that must constrain any exhibition about a pastellist. The NG could have chosen to place Liotard’s work in the context of his rivals: work by La Tour and Perronneau from the gallery’s own collection might have been included, and borrowings of pastels within London could have been chosen.
I’d also have liked to see the NG’s own Chardin: the Petite Maîtresse d’école from the 1740 salon (the theme of the conversation piece in pastel would have been a focused study, requiring input on the technical side – limitations of size – with national interests: Knapton and Cotes would be more prominent than French pastellists, while Liotard’s influence on pastellists such as Guillibaud will not be known to many visitors). In the NG painting, Chardin, the still-life painter, concentrates on the faces and the bond between the girl and her pupil in a work which is filled with humanity; Liotard, the portraitist, focuses on the accessories, and puts the aunt’s face in half-shadow. It isn’t difficult to see who wins.
The exhibition borrowed five pastels, four from outside London which have had to travel long distances (Chatsworth is 160 miles from London, Compton Verney nearly 100). Readers of this blog will know that I and many others do not consider this wise. And even if lenders are willing, one has to wonder whether it is responsible for the National Gallery to accept such loans – particularly for works all of which were seen recently in the RA show. That Journal18 review of the RA exhibiton was entitled “Pastel will travel” – so we may infer that the author does not agree with my caution – although she does record surprise that the pastel survived the journey from Lyon to London “miraculously…intact” (but one wonders whether that is the case, or whether Liotard might have had to retouch parts using his own materials which would not be detectable today); there is a particular irony in the director’s preface which praises the previous owner’s “zealous care” in looking after the picture, including refusing to lend to temporary exhibitions.
Of the loans, by far the most impressive is Lady Fawkener – perhaps the star of the whole show, just as she was in 2015; certainly a picture for which I have a particular faible. Let down only by Liotard’s inability to depict hands, it is an exquisitely beautiful example of just what Liotard could achieve working on parchment, his preferred support; and at the same time it has a depth of psychological interaction with the sitter which draws the viewer in. I’m not sure why the two Déjeuners were put on a screen in the middle of the room: there was plenty of space to put them all round the outer walls, allowing an easier comparison of Lady Fawkener with the NG pastel.
Another direction the exhibition might have taken further is the choice of medium, as I note above. The two most striking things about the confrontation of the two versions pass almost unnoticed. The first is that in choosing to reproduce his earlier pastel in oil, Liotard tells us that he no longer believes the magic is in the medium. The obsessive level of detail in reproducing chalk strokes in oil is a theme that could be developed further. His other repetition after a long interval, the second version of Mme Necker which I published from Graf Zinzendorf’s acount (see my essay Liotardiana from p. 11 on), was in the same medium; but here the use of oil, not simply to replicate the picture, but to replicate the exact pastel technique that created the picture in the first place, seems bizarre. One speculates that the pastel had already deteriorated to some degree and Liotard wanted to capture a permanent record of it (he nearly succeeded, but for the blue on the cups as the catalogue discusses); and it is of course the oil version whose perfection he discusses in his Traité.
The method of such duplication remains somehwat obscure. Care should be taken in assessing the tracing in Fig. 53 which seems to me for a number of reasons unlikely to be autograph. Indeed I persuaded Marcel Roethlisberger to change his mind on this, and in an unpublished letter to the Dresden curator in 2017 Roethlisberger wrote “Neil Jeffares verwirft Liotards Eigenhändigkeit der Zeichnung. Im Gegensatz zu der von mir im Buch geäusserten Meinung bin ich nun ebenfalls dieser Ansicht.”
The second issue is equally fundamental. It is not only the composition of the sticks that matters, but what they are applied to. For me the key issue with Le Déjeuner Lavergne is how this painter of immaculate surfaces ended up with the sticky mess we see on the flesh areas – a feature which isn’t touched on in the catalogue (and to be fair the subtle lighting in the exhibition makes it much less evident than when the pastel was originally displayed at the gallery). At first sight one is tempted to dismiss this as unfortunate later restoration, but it isn’t. Particularly for Liotard, the painter of surfaces, perfection in textures matters in a way it might not with, say, Chardin; and some of his other work sets a high standard for what he advertised as “his highest finishing” – his “point de touches” rule. The explanation for all this is surely in the choice of support. Liotard’s ambition to paint two large conversation pieces forced him to switch from parchment to paper (only later would he source larger sheets of parchment), and the techniques for working and in particular erasing worked differently. The result is that he overworked the flesh areas with more layers than he had needed when working on parchment.
The NG catalogue tells us repeatedly that this is his masterpiece (some earlier versions of the publicity still floating round still claim it erroneously as his largest ever pastel), without noting that the only source for the contemporary claims was Liotard himself. We are told that Horace Walpole counted the pastel among Liotard’s “most capital works” – but although the Anecdotes say that “the Earls of Harrington and Besborough have some of his most capital works”, the footnote directs us to the Breakfast belonging to Harrington (now in Munich, fig. 66), not this. And while Roethlisberger & Loche do call it “le chef-d’œuvre de Liotard, s’il fallait en désigner un”, I’m not sure under what conditions they were able to see it. That is not to say that it is not a very remarkable work and quintessentially Liotard, but a plea to the National Gallery to put it in a context beyond tea-trays.
Neil Jeffares
Le Pastel, entre ligne et couleur. Chefs-d’œuvre des collections du musée Cognacq-Jay, Paris, musée Cognacq-Jay, 12.x.2023 – 11.ii.2024
From now until 11 February next year you can see an interesting display of eighteenth century pastels at the musée Cognacq-Jay, comprising mainly their own holdings but supplemented with a small group of pastels from affiliated musées de la Ville de Paris (all the works below are in the Cognacq-Jay except where otherwise indicated). The organisers, Pascal Faracci and Sixtine de Saint-Léger, have judiciously chosen to restrict the exhibition to a smallish group of outstanding quality rather than dilute it with a broader gathering of whatever is available: I have written before and again about the perils of transporting pastels to temporary exhibitions, and how it can introduce a bias when the best works are not available. The quid pro quo however is that there seems to be no printed catalogue – one is thrown back to Thérèse Burollet’s 2008 catalogue, Musée Cognacq-Jay. Pastels et dessins, which though readable and attractive requires some commentary. The cartels are surprisingly brief; this might have been an opportunity to cover more ground. There is a useful, if incomplete, dossier de presse here.
I thought it would be helpful therefore to list the exhibits (but I will refrain from reproducing each of them – just go and see them!) and add some observations, specifically to note where I disagree with Burollet or to add information she wasn’t aware of; and also to highlight some questions which might be investigated further by the curators. Of course full details of each pastel (dimensions, provenance, exhibition and literature) are included in my online Dictionary of pastellists: this exhibition is coded as Paris 2023b, and if you key that into the search box on my home page you will be taken to all the exhibits. I also include their J numbers below, the simplest way to uniquely identify any pre-1800 pastel (again just key in to the search box – but remember, no spaces!), with links to the fascicle or pdf of the Dictionary in which the entries appear and are reproduced.
Sadly my health prevents me travelling at present, so I’m not reviewing the display as such, but I have previously seen all the exhibits. (I haven’t however seen each pastel out of its frame, and of course would remind you that attribution is a subjective matter and these comments are subject to the standard disclaimer on my website.) Images on social media or shared by friends (some of which I’ve reused here with thanks) do suggest I’d have something to say about the hang of some of the pictures in narrow corridors or the directional lighting but let us not pursue that. I’ve written before too of my wish to see the day when pastels could be exhibited without the need to have a display showing the materials, beautiful though the materials undoubtedly are.
J.173.464 Boucher, Étude de pied (musée Carnavalet, inv. D.4353) 1752
See fascicle. Boucher is not always thought of as a pastellist, and the position is complicated by the vast number of misdescriptions and misattributions. Among the true pastels by him, the Étude de pied in the Carnavalet is a masterpiece combining both the subtle luminosity of the pastel medium with the author’s skills as a draughtsman. Unlike the other pastels in the exhibition, it is of course a preliminary study – for his 1752 painting of a Femme nue couchée sur un sofa (supposedly Mlle O’Murphy) of which there are several versions (most notably in the Alte Pinakothek, Munich). It demonstrates the hatched, linear approach characteristic of this artist’s drawings. As such it more properly bears out the exhibition’s subtitle, “Entre ligne et couleur”, than most of the other exhibits which are more within the “peinture au pastel” bracket.
J.338.1028 Gardner, Lady Auckland and her daughter c.1777
See fascicle. A typical example of Gardner’s uniquely personal style. Is it “a pastel”? His preferred technique used dry pastel for faces and gouache liberally applied for the rest of the picture. When I started writing the Dictionary more than twenty years ago I was minded to exclude, or just have a sample of his work; but the space freedom of online publishing allows me at least to try to include all such works by this extremely prolific artist. It is incidentally pure coincidence that the Cognacq-Jay owns portraits of the wives of two successive earls of Buckinghamshire: the provenances are independent. Lady Auckland and child is in one of Gardner’s standard fluted frames. (The cartel includes a detail of a dog from the Russell pastel, I assume by mistake.)
J.338.1086 Gardner, Lady Buckinghamshire c.1775
See fascicle. Until the reemergence of a second version of this pastel (at a sale in Sotheby’s in 2022) which I had the opportunity of researching very thoroughly, the Cognacq-Jay picture was assumed to be the only version, and to have been the one in the family collection at Nocton; its authenticity was not in doubt. Unfortunately I have had to reconstruct a quite different history for the two versions than you will find in the pre-2022 literature.
It is now clear that the other version, J.338.1085, was that in Nocton Rise up to the 1914 sale. (The vendor was in fact Edward John Howard, not Arthur Grenfell who was the source of different lots in the same sale. The confusions are compounded by the unrecorded fact that Howard was the grandson of the 3rd Earl of Buckinghamshire from an illegitimate child the peer had by his housekeeper.) From a curious family history book by the sitter’s great-granddaughter, Albinia Cust Wherry, The Albinia book, London, 1929, we learn that the picture depicts Albinia Bertie as as Lady Townley in Vanburgh’s The Provok’d Husband, in an amateur theatrical production at Nocton on Boxing Day 1775. Lady Brownlow described it in a letter to her mother a few days later: “Mrs Hobbard [our sitter was then plain Mrs Hobart] was very great in Lady Townly.” (I am most grateful to a private collector who was able to supplement my research on this picture.) The portrait was very probably executed soon after.
The key point, once the provenances are unravelled, is that the Nocton picture (right) came onto the Paris art market in 1914, with Jacques Seligman, at a time when Gardner’s prices had taken off (he had previously been regarded as very minor; it was the 1908 sale of his portrait of Lady Fawkener for 1250 guineas that brought the artist out of obscurity). Soon after it was with Duveen, a dealer with a mixed reputation. The Cognacq-Jay version (left, above) however as far as I can establish was not known before Ernest Cognacq acquired it, by 1924. That in itself doesn’t prove conclusively that it was a copy – although the improbability of a second version popping up in Paris independently is the obvious concern; had the sitter commissioned a replica, perhaps to be given to a friend, it is more likely to have emerged in the London market before appearing in Paris.
We have to fall back on visual evidence and connoisseurship unless further evidence of the provenance of the Cognacq-Jay version emerges. There are numerous tiny differences, mostly in the unimportant parts of the composition, e.g. the shape of the muslin bustle behind her dress, the clouds in the top right behind the plant in the urn, the shape of the tiling on the floor in the lower right foreground, the definition of the fabric etc., not to mention the signature on the Nocton version (although Gardner in fact rarely signed). Mostly these changes are simplifications rather than imaginative variations. There is also a more significant difference: the Nocton version was squeezed into its Gardner frame by cutting corners on the right hand side, thus hiding behind the frame a vertical band of about 6 cm: this results in the somewhat odd mise-en-page where the sitter’s head is no longer centred; the Cognacq-Jay version simply copies the visible part of the Nocton composition, suggesting it was not made in the studio. Here is the full Nocton version out of its frame:
But for me the crucial difference is in the textures of the surfaces. Gardner’s style is a very broad brush gouache which is extremely easy to imitate. But he has one very personal idiosyncrasy: for the face he switches to dry pastel and graphite. There is pastel in the Cognacq-Jay face and hands, but it doesn’t seem to be applied in the same way as in the other version (see comparison, again CJ left, Nocton right):
Even details like the sky and foliage reveal the same difference in treatment which strongly suggests a different hand (same order):
I haven’t been able to examine the Cognacq-Jay version out of its frame (it appears to be mounted on board rather than marouflé sur toile as the other is), nor do I know if spectroscopy could cast some light on the pigments used. It is clear from the Nocton version, for example, where some of the “blue” foliage was hidden under the rebate, the colour was originally green – the typical eighteenth century problem of greens being a mixture of a stable blue with a vegetable dye called stil-de-grain which fades rapidly in sunlight. Is this the case with the Cognacq-Jay version, or has a uniform blue been used that matches the condition of the original c.1920? My comments here are of course subject to further information that might emerge from technical or archival research.
I have no doubt that the Nocton version is an autograph work by Gardner. I think it unlikely (but not impossible) that he produced a second version, but I am sceptical that he would have made a second version so closely following the Nocton one, but in a quite different technique.
J.375.1125 Hamilton, Lady as Emma, the Nut-brown Maid c.1779
See fascicle. Allegorical portraits of ladies posing as Emma, the Nut-brown Maid – normally from the Matthew Prior poem, but there were other versions of the old ballad – abounded in Britain throughout the eighteenth century (as Burollet notes). While the subject may have been inspired by Francis Cotes, the composition is closer to Reynolds’s Countess of Fife, 1764–65 (Mannings 527).
The dating of this is a little problematic. Almost all of Hamilton’s full length pastels were done in Italy where he spent the 1780s. There are a small number of exceptions, and I think (based on hairstyle etc.) that this must be one of them, done in Britain before his departure for Rome which is usually placed in 1779 but might be up to a year or so later. There is an interesting point of comparison (although the difference in size means it is not actually a pendant) with the only other allegorical work in his œuvre (and apparently the only female full length: but then it was almost always the males in British families that went on the Grand Tour), the pastel of the Muse Erato, no. J.375.2164 in my catalogue (below), painted before his Italian trip as Crookshank & Glin 1997 (p. 66) acknowledge, suggesting that the fashionable lady was possibly part of the Duke of Leinster’s circle. But Crookshank & Glin infer from the identification of the Cognacq-Jay sitter as “Lady Carhampton” that that pastel must have been made after 1787 (when the only likely candidate for that name became known by it, after her husband became 2nd Earl of Carhampton) and therefore in Italy.
What Crookshank & Glin have missed (and Burollet too) is that the identification is nonsense. The Cognacq-Jay sitter, like the “Erato” sitter, is unknown. She was “inconnue” when Cognacq bought it and when Ricci and Jonas catalogued it. In 1980 Burollet noted that the face resembled strongly that of a small oval Hamilton pastel published by its owner, Edward McGuire, in 1937 as of “Lady Carhampton, née La Touche”; she inferred that the Cognacq-Jay sitter was the same lady. Two problems: there was no Mlle La Touche who became a Lady Carhampton; the only possible candidate for a Lady Carhampton is Jane Boyd (c.1755–1831). She was the daughter of George Boyd, a wealthy Dublin wine merchant, bringing a dowry of £20,000, and in 1776 she married the 3rd Earl of Carhampton (whose earlier portrait by Hamilton is in the National Gallery of Ireland, J.375.1121). Burollet 2008 writes “Jeffares (op. cit. p. 223) propose une autre origine « née Jane Boyd » sans aucunement préciser ses sources”: I don’t cite references which can be verified in every standard genealogical reference work, e.g. those cited on this page of my site. But the second problem is that McGuire’s pastel has subsequently emerged, and the colour photograph (which Burollet evidently didn’t know) shows that her eyes were blue, while the Cognacq-Jay sitter’s are brown. (One might add a third problem: anyone seeking to identify sitters in Hamilton’s portraiture by resemblance alone is embarked on a hazardous mission.)
In conclusion: the Cognacq-Jay have a pastel of an unknown lady probably done c.1779 around the same time as Erato.
J.46.113 La Tour, Autoportrait au jabot 1750
I have written at considerable length about the versions of the Autoportrait au jabot both in my La Tour catalogue (pp. 68ff of the 2022 edition of my La Tour catalogue raisonné, available also in this fascicle from the up-to-date version which may be easier to download; you can find links to all the elements of my La Tour catalogue from this page) and in the specific essay. I hope the arguments are clear, but I’d prefer not to abbreviate them since only the full intricacies of the logic force us to recognise that the much more famous and widely reproduced Amiens version of this portrait must be a studio copy by La Tour’s pupil Montjoye. The Cognacq-Jay pastel is thus promoted to the primary surviving version of this Enlightenment icon.
J.46.2725 La Tour, La présidente de Rieux 1742
This magnificent pastel is of course the star of the show, and will draw people in at the door. My catalogue has a full discussion (at p. 206 of the 2022 edition of my La Tour catalogue or pp. 7ff in this fascicle).
J.46.319 La Tour, inconnu au gilet bleu c.1750
Again my catalogue has as much as can be said about the inconnu au gilet bleu (fascicle).
J.46.2852 La Tour, marquis de Sassenage ?c.1748
See fascicle. The debate here is over the identification of the sitter and whether it can be the pastel of the “comte de Sassenage” exhibited in the salon of 1748 given that Sassenage was only made chevalier du Saint-Esprit in 1749. Could the cordon bleu have been added later, as happens in some portraits? I doubt it as the ribbon and badge seem to be quite well integrated, but it is possible that scientific examination might throw some light on this conundrum.
J.46.2872 La Tour, Maurice de Saxe (musée de la Vie romantique, inv. D.8330/D.89.48)
A version of a much repeated portrait of which the primary version is in the Louvre (J.46.2865); see fascicle. It might be thought that the provenance – a gift from the sitter to his mistress, with descent in the family of the sitter’s granddaughter, George Sand (who even described it in a passage in her Histoire de ma vie: included in my Florilegium), until given to the museum in 1923 – should guarantee authenticity, but there are a number of contemporary copies of the La Tour masterpiece in the Louvre, and a close comparison seems to me to show that the copyist here followed the original more slavishly than La Tour himself would have done in an autograph repetition, where the approach would have been to recreate effects rather than imitate strokes. To take a single example, look at the pair of double white strokes on the sitter’s upper lip, and contrast the life in the Louvre version (right) with the pedestrian reproduction in the Vie romantique version:
J.9.2709 École française, la marquise de Sassenage c.1762
See fascicle. The Cognacq-Jay pastel, in my opinion, is not by La Tour, but most likely one of a number of copies of a portrait of which I am unconvinced that the original has been identified. (For that matter, neither has the sitter been identifed securely.) More to the point: is the original by La Tour or another artist? We can only judge by the pose, not the technique. To me the full-on orientation of the body and the angle of the head are closer to Nattier (but the technique certainly far short of his), and perhaps the work is by one of his followers (Mérelle perhaps?). Absent the original (not necessarily a pastel) this is all guesswork, which is why I prefer to catalogue this as École francaise rather than “after La Tour”. It is curious that several other versions exist; one was with baron Emmanuel Leonino, while another, now in Sassenage, was commissioned in 1913 by Édouard Jonas for Pierre-Luc-Raymond, marquis de Bérenger, in part exchange for the version now in the Cognacq-Jay. This is clear from a letter from Bérenger to Jonas in the archives of the château de Sassenage (now delightfully online) setting out the terms of the sale, and including this clause:
Vous aurez à me fournir gratuitement et à vos frais une copie exacte et bien faite de chacun des tableaux et pastels mentionnés dans la vente ci-dessus destinés à remplacer ceux vendus. Vous aurez également à me fournir gratuitement et à vos frais des cadres reproduction exacte des quatre cadres des pastels. Les copies des tableaux et des cadres des pastels devant m’etre remis dans les trois mois de en date la vente […]
This practice of dealers (Duveen, the Wertheimers etc.) inducing reluctant descendants to part with original 18th century family pastels was widespread at the start of the twentieth century. We don’t know the names of the copyists but some were of astonishing technical competence and frequently continue to confuse.
J.478.173 Lenoir, Comus (musée Carnavalet D.4361) 1776
See fascicle. The inclusion of the portrait of the magician and scientist is a reminder of the depth of talent in the French school: while Rosalba and Liotard were isolated sparks of genius, behind La Tour and Perronneau was a whole school of outstanding artists, from Vivien to Coypel, Loir, Valade, Bernard, Hoin, Suzanne Roslin, Labille-Guiard, Boze, Ducreux and so many more – as well as painters who occasionally took up pastel: Nattier, Chardin, Greuze, Fragonard etc.
There is in Lenoir’s apparently run-of-the-mill portrait more than might be immediately obvious. When it was exhibited at the 1779 salon (together with two others), the critic in the Mémoires secrets commented on Lenoir, a newcomer to the official salons (he was previously at the Académie de Saint-Luc):
Un Agréé débute dans le genre de MM. Duplessis & Roslin, mais non en imitateur servile: ses portraits de M. Morand, de M. de la Blancherie, de M. Comus font déjà plaisir & sont variés comme les personnages que rend son pinceau. Le Docteur en médecine, à travers la magnificence de son vêtement, a la gravité qu’il doit avoir, & l’esprit qui est dans ses yeux est celui de son état, un esprit réfléchi & profond. Le caractere juif, empreinte de la figure du second, est saillant; & la gaîté fine de l’escamoteur brille sur sa face fleurie: entre ces deux charlatans le spectateur se sent disposé à rire d’être dupe de celui-ci; il seroit fâche de l’être de celui-là, dont la mine pédantesque trahit la nullité sous un air scientifique. M. le Noir, c’est le nom de l’artiste, se signale ainsi entre ses confreres, par sa finesse pour exprimer les pensées sur les physionomies: il ne sauroit trop cultiver cette qualité, la plus précieuse et la plus difficile du genre.
J.582.1318 Perronneau, Mme d’Espremenil c.1768
See fascicle. The question here is over the identification of the sitter which I believe is based only on an old label (which I haven’t seen, and can’t therefore assess if it was written when the portrait was executed) identifying her as “Mme d’Epremesnil” (various spellings; the differences are not relevant). It has been assumed this must be Françoise-Augustine, dite Éléonore, Sentuary (1749–1794), who at the time of the pastel (which must be in or close to 1768) was Mme Jacques Thilorier, and only later became the second wife of Jean-Jacques Duval d’Esprémesnil. But can we be sure the portrait is not of his first wife (the only “Mme d’Epresmenil” in 1768), née Marie-Madeleine des Vaulx d’Oinville (1747–1774)? There are several arguments against this (otherwise more logical) identification. The first is that Perronneau was in Bordeaux in 1768: but not for the whole year, so this doesn’t rule out a Parisian for a portrait “vers 1768”. Secondly the proposed link with another portrait, no. 302 in Arnoult (my J.582.1369), which she bases on the rather dangerous argument that both sitters show their mantles fixed with a similar clasp. The dimensions of the pastels are quite different. We know that a portrait of Mme Guesnon de Bonneuil was exhibited in the salon de 1769, which d’Arnoult assumes is by Perronneau – although Bachaumont does not name the artist or identify the medium: but consulting the livret of the Salon one finds that this reference (including the spelling mistake, Ponneuil for Bonneuil) is to an oil portrait by Alexandre Roslin (no. 42) (note that Bachaumont also includes in this list abbé Jourdans, a painting by Duplessis, no. 197 in the livret, so there is no implication that he refers only to Perronneau sitters). Finally there is a much later oil portrait (reproduced in Olivier Blanc’s Portraits de femmes, p. 281), allegedly by Vigée Le Brun (although Joseph Baillio agrees with me that this is not correct) of a lady said to be Françoise-Augustine Sentuary and bearing a vague resemblance to the Cognacq-Jay sitter: but this is not sound evidence. There is thus nothing to connect Perronneau with either of the Sentuary sisters, and I think the identification should be treated with caution, the first wife logically being more probable.
J.582.1536 Perronneau, Le Normant du Coudray 1766
See fascicle and my essay Jeffares 2015n “The fixer fixed: Loriot in Orléans”, 24.ii.2015.
J.64.2219 Russell, Mrs O’Shee 1789
See fascicle; a splendid example of Russell’s art. Although difficult to read, the signature is probably “J Russell R.A. pinxt/1789” rather than as reported in Burollet.
The identification is probably correct (if understandably misspelled: not O’Shea, but O’Shee, the rarer spelling) although a little surprising in view of the apparent age of the sitter (ask yourself before reading on…but it is easy to see why she was labelled “Miss Power, plus tard Mrs O’Shea” by Burollet, who thought she was a “jeune fille”). Burollet’s authority for the name was Francis Webb’s Russell album (V&A), in which the annotation to the photograph is actually “Miss Power of Garden Morris Waterford. Cr 1789. Married Mr Shea of Sheadown (later O’Shea).” Webb isn’t always strictly accurate, but the “later” refers to the adoption of the prefix O’ by her husband, not to the marriage.
Unfortunately Irish genealogy records suffered a similar fate to Paris parish records, in 1922 instead of 1871, so there are gaps, but I have spent quite a lot of time unravelling this in case there was a confusion with her daughter-in-law who had a similar name. The Cognacq-Jay sitter must however be Mrs John O’Shee, née Elizabeth Power (Gardenmorris, Waterford a.1748–p.1811), daughter and heir of Richard Power of Garden Morres, Co. Waterford, who on 23.ix.1767 married John Shee, of Sheestown, Co. Kilkenny, later O’Shee; he was High Sheriff for the county of Waterford in 1783 (his brother and a number of close relatives were officers in the French army). The union brought together two families of Irish landed gentry, hers staunchly Protestant, his recently converted. She was born at some date before 1748 (one source states confidently 1740), and so was at least 41 at the time of Russell’s pastel – and had been married for some 21 years. Her daughter-in-law, Margaret Elizabeth Power, who married her son Richard Power O’Shee in 1804, was not born until after her own parents, Nicholas Power of Snowhill, Kilkenny and Rachel O’Neill, were married in 1780, and cannot be the adult depicted in the pastel.
Records don’t allow us to confirm Mrs O’Shee’s visit to Russell’s studio in London, but she is likely to be the Mrs O’Shee mentioned among six English ladies of fashion sporting the latest white stockings in the German Spa in Liège reported in The World on 24.vi.1788. Perhaps she passed through London after a Continental tour.
Although of no relevance to the Russell pastel, readers might be curious to see this Irish 19th century imitation of French architecture at Gardenmorris. The IRA burned it down during the Troubles.
J.76.173 Vigée Le Brun, la princesse Czartoryska (Petit Palais, inv. D. Dut. 1168) 1801
See fascicle. With this study by Vigée Le Brun we are already in the nineteenth century. It is the other picture in the exhibition that is preparatory to an oil painting (in Warsaw, executed some time later from the pastel made in St Petersburg). Although the artist did these free drawings with pastel from well before this date, we can see the world of difference opening up with the mid-century peintures au pastel of La Tour and Perronneau. I can refer you to the entry in the outstanding catalogue for the 2015 Vigée Le Brun exhibition (cat. no. 134 in the French edition).
Van Dyck, Nicholas Lanier (Vienna, KHM)
Art historians spend a great deal of time poring over old sales catalogues in the hope of recreating provenances, and can draw on vast resources such as the Getty Provenance Index or Lugt’s Repertoire des ventes (and, for those who can afford the astronomical subscription, the commercial Art sales catalogues online database). But sometimes the missing clues are to be found in genealogies or wills – documents which are also increasingly available online, but seldom indexed for the benefit of those searching objects rather than people. Since discoveries thus remain a little aleatory, I thought I would report one which isn’t in my normal field but will be of interest to those researching Stuart collecting. I leave it to those scholars to pick up the hints this document presents.
It relates to a certain “William Towers” who is sighted several times in relation to Samuel Cooper and his circle, including members of the Gibson and Hoskins families of miniaturists. Interest was triggered initially by a group of drawings attributed to Cooper and Richard Gibson shown in a Royal Academy exhibition in 1960, The age of Charles II. They belonged to a collector, a Mr A. H. Harford, who had an ancestor called Thomas Tower [sic] of Tatham in Lancashire, who died in 1659. Mentioned by Daphne Foskett in her 1974 monogaph on Samuel Cooper (p. 85), the trail was picked up by Mary Edmond in a long study of “Limners and picturemakers” (Walpole Society, xlvii, 1978, pp. 60–242, at p. 110): she noted that there was a Mr William Towers who was known to Richard Gibson – as well as to Mrs Elizabeth Lanier, and plausibly (but as we shall see erroneously) suggested that this William Towers must have been an ancestor of Mr Harford, the Towers/Tower names being so similar.
Elizabeth Lanier was the widow of Nicholas Lanier (1588-1666), the musician whose portrait by Van Dyck once belonged to Charles I and is now to be found in the Kunsthistorisches Museum in Vienna (as I have no image of Towers, that masterpiece adorns this blogpost: whether it is the picture mentioned below is a question I shall leave to specialists. Lanier bought his portrait at the 1649 sale of the king’s collection, and there is a gap in the provenance before it appears in Vienna in 1720). In her will, made 1672 and proved the following year, she gave “unto Mr William Towers Two of my best Pictures” as well as “two more of my Pictures and forty shillings in money” to Mr Richard Gibson (perhaps he needed the cash more that Towers).
When Gibson wrote his will in 1677, although of course it was not proved until 1690, he too mentioned his “worthy friend Mr William Towers” and gave him his “Little Birding Gun made by Harman Barne” (a well-known gunmaker, appointed to the king and to Prince Rupert).
Armed with the conviction that she had the name of the early owner of these sheets, Edmond set about to identify this William Towers more specifically. He must have been still living at the time of Gibson’s will (easily thought of as 1690, although of course 1677 is the right terminus post quem to infer), and it would help if he lived in the same parish, St Martin-in-the-Fields. Edmond found a plausible candidate in a William Towers, freeman of the Feltmakers’ Company (so I shall call him here the feltmaker), who had died in 1693, and tentatively suggested him. Her research established that the feltmaker’s probable grandfather came from Lancashire, as did Mr Harford’s ancestors; but she seems not to have noticed that the feltmaker’s son died the following year, making a will covering also his father’s estate, and bequeathing everything to a Miles Spensley, a cooper, who had no obvious connection with the Tower/Harford line; nor was there any evidence of an interest in art or a social status that would have made that more plausible (particularly in the light of the information below).
By the time the drawings were shown in the British Museum show Drawing in England from Hilliard to Hogarth, 1987, with an otherwise excellent catalogue by Lindsay Stainton and Christopher White (nos. 63, 64, 75, 76 and 80), the owner’s line back to the feltmaker Towers (d.1693) was considered certain and has since been repeated in other sources.
A few further facts about Gibson’s friend were gleaned by other scholars, notably in an article on Nicholas Lanier by Jeremy Wood (Collecting prints and drawings in Europe, c. 1500–1750, ed. Christopher Baker & al., p. 120) where the William Towers mentioned in Mrs Lanier’s will is identified as the former keeper of the Earl of Pembroke’s collections. The reference cited is an article by Philip McEvansoyea (“The sequestration and dispersal of the Buckingham collection”, Journal of the history of collections, viii/2, 1996, at p. 135), according to which a Mr Towers (or Touars) was being paid £20 a year by the Earl of Pembroke to look after the pictures at Durham House and at Baynard’s Castle. This was derived in turn from the 1968 catalogue of the Wilton pictures, now superseded by Francis Russell’s account (The pictures and drawings at Wilton House. 2021, at p. 5), where a further reference to Towers in Aubrey’s Natural history of Wiltshire (1847 ed., p. 91) is also mentioned. That passage in Aubrey is worth citing in full as it explains the link between Towers and Gibson:
Russell goes on (p. 6) to list several transactions which Towers executed in connection with the winding up of the 4th Earl’s estate: in September 1651 he paid £535 “for several pictures sold by him to one Monr Tenier a Dutchman”; in August he sold three pictures to Lord Bellasis fror £40, and later that year sold pictures to the value of £213 to the Spanish Ambassador, Alonso de Cardénas, of which £4 was to cover a gratuity to “one that helped in the sale”. Sales continued until 1654, the last being “a flower pot done by Clare a wooman” [Clara Peters].
There is one further sighting of Towers in the Pembroke literature – in the 1731 Description of the Earl of Pembroke’s pictures by conte Carlo Gambarini. The volume is cited in Russell’s general references, but omitted from the specific literature given for what we can agree is the “most spectacular and compositionally ambitious of all van Dyck’s portrait groups”, viz. The Pembroke Family.
Van Dyck, The Pembroke Family, Wilton House
Russell lists a later engraving by Bernard Baron (1740 – after Gambarini was published), but not the Audran, nor (as far as I can see) is the involvement of Towers discussed. The reduced version in the Hermitage (link), at one time attributed to Lely but now catalogued as by an unknown hand, is presumably the copy Gambarini mentions as made by “Remy” specifically to be engraved and then owned by Pierre Crozat. (This must I think be Remigius Van Leemput (1607-1675); while in Crozat’s collection, where it was seen by Richardson in 1716, it was confused with the English royal family.) Here anyway is the passage:
Time now to reveal the true William Towers who looked after the Earl of Pembroke’s collection, befriended Elizabeth Lanier and Richard Gibson and collected paintings. The will is available online (from Ancestry if you have a subscription, or from the National Archives if you do not), but as these are the proved file copies rather than autograph documents it will be easier if I give my transcription. I have added line breaks and abbreviated the monetary amounts given both in words and figures but otherwise preserved spelling; my annotations are given in square brackets.
Will of William Towers of St Martins in the Fields, made and proved 1678:
In The name of God Amen.
I William Towers of the parish of St Martins in the Fields being infirme in body but of perfect mind and memory Thanks be to God Doe make and ordaine this my last will and testament First I give and bequeath my soule into the hands of Almighty God firmly trusting to be saved by the meritts of the death and passion of my Saviour Jesus Christ And my body I comitt to the Earth to be decently buried att the discretion of my Executor hereafter named And as far and concerning that Temporall Estate wherewith it hath pleased God to blesse me I give and dispose the same in manner following (that is to say) I give and devise
Inprimis to my deare sister Mrs Anne Stelling at Levingate Read Marshall in the Bishopprick of Durham and to the helpe of her Children the sum of £40 [Clement Stelling was parish clerk in Redmarshall, Durham]
It. To the widdow of Thomas Towers lately deceased in Windsor for her children £20
It. To the sister of the said Thomas Towers a Bakers wife neare the Castlegate for herself and her Children £20
It. To my sister Francis Towers £25
It. To George Wood £5
It. To Wm my servant over and besides what shall be due unto him for his wages the sume of £10 with such of my apparell as he shall order
It. To my Landlady £5
It. To her maid servant forty shillings
It. To Mrs Lannier £5
It. To my Laundress £3
It. To Mr Banks forty shillings
It. To the poore Beggar woman with Crutches that to comes to this dore forty shillings
It. to the two [?]victims and Mr Adelham to each forty shillings (£6)
It. to be distributed amongst other poor PP Ten pounds to pray for my soule
It. to my very good friend Mr Cholmley [probably John Cholmley who died 1693 leaving a ring to the Earl of Pembroke etc.] to whom I have always been obliged and ever found a very just man, my little Diamond Ring and the picture of Doctor Faustus
It. to my Landlords sonne twenty shillings
It. to Henry the Apprentice in the Shopp tenn shillings
It. Twenty pounds to be distributed among the poorest and most distressed persons in generall
It. to my serving maid Dennis £5
It. to Mrs Gowe forty shillings to buy her a Ring
It. to Mr Jo. Hodleston the Picture of St Mary Magdalen of Guido and £5
It. The great Picture of Christ at Supper with his disciples in the Castle of Emmaus To the RR Fathers of Somersett House for their Refectory humbly begging the Charity to be buried there by them
It. Mr Gibson [Richard Gibson] my Porfery Grinding Stone and the Picture of the old Earle of Pembrook in the Closet.
It. to his daughter Mrs Rose [Susanna Penelope Rosse] the Picture of Sir John Thoroughgood [Sir John Thorowgood (1588–1657), MP, secretary of protégé of the Earl of Pembroke] in the Closet.
It. to Mrs Bawtrey the Picture of Mr Lannier [perhaps the Van Dyck or a version of it] and the little picture of Snellings [the artist Matthew Snelling done 1644, and later in the Rosse sale 1723: see Foskett p. 77] done by Couper
It. to my Landlord my Great Cabinett.
And as for and touching all the Rest of my Estate Reall and personall of what Nature kinde or quality soever I give devise and bequeath the same to my loving son Francis Towers And I doe hereby Constitute and appoint my said son Francis Towers sole Eexcutor of this my last will and testament In witness whereof I have hereunto put my hand and seale this sixt day of May in the thirtieth yeare of the Reigne of our Soveraigne Lord Charles the Second by the grace of God King of England Scotland France and Ireland defender of the Faith Anno Dno 1678. /Will. Towers/Signed sealed and published in the presence of Ge. Sayer/James Bagnall/John Eason
Probate was granted to “Francisci Towers filij dict defunt” etc., on 6 June 1678.
Although the will contains some hints about his family, they are not easy to trace to reconstruct a full family tree: for example I can find no document mentioning Towers’s son Francis. But a conveyancing document (published in the Proceedings of the Society of Antiquaries of Newcastle upon Tyne, 1921, p. 169) includes vital information:
1664, Jan 13. Release, between (1) Thomas Towers of Windsor, Surrey, gent. (only son and heir of Thomas Towers, late of St. Martins fields, Middlesex, gent., decd., which said Thomas was eldest son and heir of Thomas Towers, sometimes of Girsby, and late of Darlington, gent.) and Anne Towers, only sister of the first named Thomas; and (2) William Towers of St Martins in the field, gent. (one of the sons of Thomas Towers, late of Darlington)
In other words our William Towers who died 1678 was the younger son of Thomas Towers of Girsby, near Darlington (in the north east of England, not the north west of Mr Harford’s ancestors): father and elder brother were both dead by 1664. Since William was active collecting for the 4th Earl of Pembroke in the 1640s, he was probably born before 1620. It is also notable that he owned a porphyry grinding stone, suggesting that he was himself an amateur artist – and surely an informed connoisseur. There is however nothing to suggest that he owned any drawings.
The will was made and proved in 1678, so he cannot be the feltmaker. Further since he died well before Richard Gibson, the theory that the Harford drawings passed to him after Gibson died is also unsound. As it turns out Mr Harford had other art collectors in his pedigree, and it is perfectly possible that the drawings were purchased after Richard Gibson’s death by, for example, his great-times-6-uncle, Thomas Walker of Wimbledon Heath (1664–1748), who owned pictures by Van Dyck, Watteau etc. – but that is another story.
We should learn by the end of this week whether the National Portrait Gallery has been successful in its attempt to save Sir Joshua Reynolds’s painting of Omai for the nation. We will all rejoice if some philanthropist has come forward to write out a cheque, but it may not be as simple as that. Indeed the whole saga has raised so many different (and apparently unrelated) questions that I thought it might be worth putting them together in a post.
There’s no need for me to rehearse the history of the painting or explain its importance: there’s a vast literature. Starting points might be the documents on the Arts Council website (e.g. this summary). One aspect of the case is that it illustrates the shortcomings of Britain’s export licence scheme. Some of these are being addressed, notably the right of owners to reject matching offers, making the efforts required to raise funding not merely Herculean but Sisyphean; and I welcome the suggestion that the Culture Secretary is considering whether the destination to which the work is to be exported should be taken into account (a work in the Louvre is easier for most of us to see than one in a secret bunker in the UK).
The editorial in this month’s Burlington Magazine (here: free access) goes further, leaning towards the conclusion that the whole system be scrapped, and museums encouraged to purchase directly from families who want to bring their heirlooms to market. I’m sure people will be ready to point out that it takes UK museums time to put the funds together to bid at auction, and with even less certainty of success than under the former right-to-reject regime, it will be impracticable to do so.
Of course the real problem is that our museums don’t have adequate acquisition budgets, UK culture does not encourage US-style philanthropy and our governments won’t simply write out cheques for old master paintings (I don’t think they’ve done so for half a century or more). What we have instead is a somewhat Byzantine scheme of tax “douceurs” that sits alongside the export licensing scheme under the aegis of the Arts Council; it survives largely because people don’t understand it. In essence family heirlooms can be retained across generations without paying inheritance tax provided certain minimal public access conditions are met; when they are finally sold, they can be accepted in lieu of tax if UK museums want them. In other words UK taxpayers are being made to do precisely what the Government refuses to do openly (since voters would object: elitism as the polar opposite of levelling up etc.). But the pictures we get this way wouldn’t necessarily be the ones museums would select if they had a free choice (so there is a steady parade of Reynolds portraits and other pictures reflecting the taste of the owners of stately homes when Britain was a wealthy country, but rarely anything to fill the gaps, for example, in French dix-huitième painting). (There is a curious tension between the desire to save “our” pictures for the nation and the argument that our museums should retain the Elgin marbles etc. as centres of world culture.)
In the case of Omai the system failed at this first hurdle. When the Howard family decided to sell the picture in 2001 (to fund a divorce), Tate raised what might have been a sufficient offer when the tax douceur was taken into account only to find that the family had a different tax plan. This depended on arguing that the picture was in fact “plant and machinery” required in the business (of attracting paying visitors to the house), and as such was a “wasting asset” and thus exempt from capital gains tax. You might think this a rather ambitious claim (although less so if you understand the technical meaning of these terms in tax law); and the Revenue did too. Their interpretation succeeded at the First-tier Tribunal, but on appeal the Upper Tribunal allowed the claim. Incidentally if you’re American, and have watched Brideshead Revisited, you probably think that Castle Howard belonged to a Lord; if you’re British you know it belonged to an Hon. – but you probably think the painting belonged to him, while in fact the painting was sold by (and the tax case was in the name of) the Executors of Lord Howard of Henderskelfe. The decision is also worth reading.
You will all have followed the subsequent (and uniquely long running) story of export licence deferral. The particular problem here is that the owner doesn’t seem to be selling the picture at all: he merely wishes to take it out of the UK (one imagines to Ireland, but we have no idea where or whether it will then be lent to a public institution). (Incidentally for a picture that has been in private ownership it has been exhibited outside Castle Howard quite regularly – in at least eight public shows during my lifetime.) So the mechanism for export licensing, which is designed to come up with a matching offer (where it cannot matter to the vendor which they choose), again works imperfectly: this time there is no hard evidence that the value proposed is real. Instead the Reviewing Committee drafted in an independent expert who, the paper I cited above reported, “agreed with the applicant’s valuation of £50m and their detailed justification of it. This was … submitted to the Secretary of State who agreed that as the fair market price for the painting.”
But all of us involved in any way in the art market know just how difficult the valuation of paintings can be. Countless pictures come to auction with estimates which are exceeded by a factor of ten, while others fail, are re-presented repeatedly and are eventually sold for a fraction of the initial estimate. The phrase in physics for this type of assessment is “not even wrong”. And that’s the point: no one has a clue (unless they know a specific purchaser with a cheque book at the ready) whether this picture will sell for five times what it fetched 20 years ago, itself I believe a world record for Reynolds which still stands.
I’m not going to derail this blog by listing the hospitals or nurses we could pay for with the money – considerations which of course are only relevant to the extent that public money is used to fund the purchase. To some of us £50m sounds like the price of a good Rubens, and Reynolds, who may well be Britain’s best eighteenth-century portraitist, simply isn’t in that league. It feels like tulip territory – negative elasticity in the language of economics, where the price builds in a component of extra value derived from the price already being supreme: a kind of feedback loop that leads to prices that bear no relation to value. You get a similar result when two agents are sent to an auction with unlimited commission bids for a teddy bear which ends up reaching £50,000.
The price is so high a hurdle that there are rumours of a time-share with the Getty: this seems to me to offer a poor solution. I think we should only save pictures for the nation which are going to be part of the “permanent collection” of the institution bidding. For me that means that every time I go there I can be sure of seeing it (ideally in the same place: I hate it when museums undertake rehangs). Whims aside, there is a real danger if pictures as big as this have to be moved repeatedly, whatever the condition report says (and what do we do if a future report says it can’t be moved again?). The costs of transportation and the downtime of being in transit mean that we pay more than half the cost of the picture for less than half the walltime.[1]
One of the arguments widely put out is that this is the best picture Reynolds ever painted. It may well be; but it wasn’t always seen as such. Thus the price it fetched at the sale of Reynolds’s studio pictures in 1796, 100 guineas, was matched by two other lots, and comfortably exceeded by three more. The top lot, Hope Nursing Love, reached 150 guineas. It is now in Port Eliot. It was accepted in lieu in 2006 with 22 other paintings (13 by Reynolds) which together discharged…just £2.2 million tax. No need to write to me to tell me that Omai is the better picture: I don’t disagree, but my point is that we didn’t always think so.
You may say that the tastes of 1796 are now irrelevant. But differences of view about the relative merits of particular Reynolds portraits have continued to the present century. It is difficult to ignore the fact that the foremost Reynolds scholar of our day, David Mannings, given the opportunity to view Omai in the context of a monographic exhibition of Reynolds portraits in 2005, wrote this in his review in the Burlington Magazine:
Miss Crewe appears to be on long term loan to Tate: I assume that means that in due course we are going to be asked to save that for the nation as well. What will that cost?
The answer today is probably a great deal less than Omai, because she no longer ticks the boxes. How much has changed in taste and fashion over just 18 years. Again please don’t bother to tell me you prefer Omai: my point is precisely that museum purchases with public money need to reflect long-term value based on aesthetic criteria, not temporary whims.
I want finally to turn to one aspect of the picture which I find slightly uncomfortable but where I know none of you will agree. All I can say is that if I’d told you ten years ago that the people of Bristol would rise en masse and tear down a statue of their local benefactor and throw it into the river you’d have thought I was exaggerating.
Omai was portrayed by a number of artists during his stay in England. There is a group portrait by Parry already in the NPG – or rather sometimes there, as it too is on a timeshare with Cardiff and Whitby (so who knows how often its appearance in St Martin’s Place will coincide with the Reynolds), and a number of others including Nathaniel Dance. There is even a study by Reynolds himself, which fetched a mere 7 guineas in the Reynolds sale in 1796, probably the one now in Yale.
These are remarkably consistent, and show us the real Mai. The Castle Howard picture in contrast displays an imaginary figure cooked up by Sir Joshua from his imagination, revealing not the person before him, but a confection in an outfit from the dressing-up box that is more Lawrence of Arabia than Polynesian. The question it raises (for me at least) is whether this is acceptable. Of course there is a long tradition of historiated portraiture, of ladies dressed (or partly dressed) as goddesses with which I have no problem, nor do I worry when Byron dresses up as an Albanian for Thomas Phillips: but here did Mai really consent to his being depicted thus? Ethnic portraits were not that unusual at the time, but Reynolds was aware that an accurate depiction wouldn’t sell so well. He was right: fantasy trumps authenticity. But will we agree in another 18 years?
[1] There is a further technical issue with such an offer as I read the rules in the Arts Council notice (2018/1: I’m not an expert, and the rules may be have some flexibility; but I suspect the owner’s lawyers will wish to see them applied consistently): a joint offer of this kind is not by a “public body”, and as an offer from a private source, the Secretary of State can only take it into account if it offers public access in a UK museum, the guideline being a minimum of 100 days a year. That would seem to require annual travel.
Postscript (11 March 2023)
Apparently now the deadline for donations has been extended to 11 April https://artfund.org/pages/help-save-joshua-reynolds-portrait-of-omai , although I can find no further confirmation on the Arts Council, NPG, DCMS or Artfund websites. This is bizarre given how much public interest there is. And once again it reinforces the idea that the rules are made up as we go along, which hardly strengthens London’s case as a world art market centre.
I shouldn’t think the owner will be happy about the further extension. But, according to the rules, “additional deferral periods are rare and normally only granted where there is a reasonably certain prospect of raising the residual sum within a prescribed timescale.” So, one infers, it is reasonably certain that the NPG will make a full offer.
Today’s Telegraph overenthusiastically prints a photo of the painting with the caption “…after a deal prevented it going abroad permanently”, the main story however merely reporting the Getty joint bid was a “step nearer.”
Postscript (21 March 2023)
Finally, an announcement from the Art Fund: https://www.artfund.org/news/government-extends-export-bar-on-joshua-reynolds-portrait-of-omai. The deferral period has been extended to 10 June, despite the fact that funds raised so far remain under half the target required.
Postscript (31 March 2023)
The widely heralded joint venture between the NPG and the Getty has just been announced. I haven’t seen the details, but it does involve regular travel for this Flying Dutchman, including, after he NPG’s reopening in June this year, a tour of the UK regions.
…to rephrase Linda Nochlin’s famous war cry from 1971. In fairness to the latest, which provokes these personal observations, A Revolution on Canvas: The Rise of Women Artists in Britain and France, 1760–1830 (by Paris Spies-Gans: Yale, 2022) is a sober, scholarly account which deserves its place in any art library, and would prompt a full review if only I were qualified to offer one – but I should confess now that it is mostly about oil painting, and straddles a period only one half of which interests me (1800 marks a shift in aesthetics every bit as marked as the register break on a piano). But I have no doubt that it will sell well and attract universal praise for its many achievements. One of which of course is the treasury of lavish illustrations wrapped in the superb design which Yale have provided – the latest in a long line of beautiful productions for which Gillian Malpass deserves our continuing gratitude.
But let’s stop at the picture on the cover. It’s an ideal choice to tell you what the book is about, and if you like it you’ll certainly enjoy the book. But is it, to get straight to the point, any good? In the two discussions (pp. 110ff, 130ff) the author sets Maria Cosway’s depiction of the Duchess of Devonshire in context, reports contemporary praise (“unanimously commended”, although Walpole’s succinct comment – “extravagant” – might not be regarded so positively: the OED has “fantastically absurd”), notes that the picture “advertised bold visual ambitions”, and states that “the canvas reaffirmed Cosway’s technical skill, while demonstrating how a female artist could effectively blend the allegorical and literary genres with contemporary portraiture to compose a highly original scene in the manner of Reynolds’s grand style.” Yes, but: again, is it any good? You’ll have realised by now that I think it’s simply dreadful.
There’s little point in my trying to tell you why I think that. But it is the fundamental question that bedevils art history. Whether we compile dictionaries of artists or write academic studies of movements, we are faced with the problem that most art isn’t great. And it’s not because it was made by women, poor, disabled or other minorities – the art that was commisioned, or exhibited, or just produced for the enjoyment of the artist alone is mostly unlikely to send shivers down the spines of today’s viewers. Most of us became interested in art because we experienced a profound emotional response to a great work, be it by Rembrandt, Rubens or Raphael.
Three famous paintings by men happen to be reproduced in the book (Gainsborough, David and Zoffany): two are incontestibly great by any standard, the third is of huge importance and interest. But how many of the other 166 works reproduced are in this class? I counted 13: almost all by French artists. (Of course a purely subjective choice. I didn’t include the Vigée Le Brun self-portrait in the National Gallery since it is but a pale imitation of the original version. Incidentally, if you want to give that artist a single forename, it should be Louise, not Élisabeth. And her stepfather’s name was Le Sèvre, not Silvestre – p. 61.) No doubt you will have made other choices, but the exercise is important if you believe that art history is worth studying because of the pictures.
So the question that I often ask (as I spend so much of my time with bad artists of either sex) is why we research the others? For my own work I can give a simple enough answer: establishing an œuvre and accurate biography for each artist is an essential platform for attribution; it’s a basic duty of scholarship. But feminist art histories often seek to draw deeper truths about the plight of artists who laboured under difficulties specific to their sex: this to me is more sociology than art history. Perhaps I felt this more as I turned the pages of Paris Spies-Gans’s book, revealing histogram after histogram plotting percentages of women appearing in exhibitions and so on. (There are some 29 charts and tables in the book, in addition to the 169 conventional reproductions.) There is a danger that this descends into what in finance is called “elevator analysis” – this went up, that went down etc. A picture may be worth a thousand words, but I don’t think a bar chart is worth so much.
And there is a specific hazard from which almost all academic art history suffers of paying too much attention to public exhibitions and critiques – although if people rob banks because that’s where the money is, art historians have little choice but to study the information available.
What is lost in the relentless focus on exhibition livrets are the artists who made their way through private commissions and lower profile networking – often far more important in portraiture (the bread and butter of eighteenth century art) than in history painting, typically conducted in larger, male-dominated studios embedded into the academic structure through the multiple armatures of genre hierarchies, criticism etc., and perpetuated today by universities’ focus on publishing, where narrative subjects are far easier to analyse and discuss at length than portraiture where the vocabulary runs dry too soon. So it is in a way an unfortunate handicap to impose on a study about women’s art. So too (although I would say this wouldn’t I?) is the reluctance to discuss minor media – miniature, pastel – where (as I have argued in many places: see e.g. §ix.3 of my Prolegomena) the barriers for women were lower.
The Académie de Saint-Luc gets curiously short shrift. Note 78 on p. 316 (citing the Almanach historique of the abbé Le Brun whom the author names as Jean-Baptiste-Pierre Le Brun, although I identified him as Jean-François Brun in a post on this blog) tells us this is because the author has chosen to study solely women who exhibited their art publicly – but of course this is my point. On page 38 we are told the Académie was “founded in 1391 and established as a teaching institition under this name in 1738”: that latter date is surely an error or misprint in the source cited (it would divert me too far to discuss which much earlier date can fill that space), but the relevant point is that the Académie de Saint-Luc provided legal cover for artists to work on commercial terms – and that it had admission procedures for the wives, widows and daughters of male members which are highly relevant to the rise of women artists in the period the author studies.
Another surprise, the lacuna however felt like the gravitational pull of an invisible planet, is the almost complete omission of Rosalba Carriera. (One of the two passing references, a quotation from a letter of Katherine Read, indirectly marks her significance.) Perhaps this is refreshing, in that far too much has been written about her already (see my review article in the August issue of the Burlington Magazine). But I find it puzzling that the most famous woman artist of the eighteenth century, arguably ever, who shaped so many of the concepts around which this book is built, is virtually omitted: her visit to Paris was forty years before the 1760–1830 period, but its reverberations are still felt today.
I did feel too, within the book’s parameters, that more could have been done to fill out the lives of the very many women whose names are preserved in these exhibition catalogues and livrets. You might say that if their work is unknown or bad this is pointless, but I do think they deserve the respect of getting their basic details correct. And I think the scholars who have unearthed new material deserve to be encouraged to continue by having such research noted (the trail is also important to avoid the propagation of single-source errors). And when we do find such material, it can offer a real contribution to the sociological aspect – as when we find that an unknown artist was in fact closely related to another, or resided among collectors, or lived far longer than her exhibiting floruit dates (so that she probably abandoned art).
Of course I’ve already done a lot of this work for pastellists. It is thus a little disappointing to find, for example, the third paragraph on p. 292. Mlle Frémy’s dates are given as 1754-1788: I thought I was the first to discover the year of her birth, which I published in the Dictionary three years ago; while I give her death as “p.1788”, i.e. in or after 1788 (I suggest she may have still been alive in 1797, but not as an artist), no reference is given for the exact 1788 printed. Similarly I published Marie-Victoire Davril’s dates in 2016 and wasn’t aware that others had found them in 2016 (they are copied in Wikipedia, but citing me as source; neither woman’s dates are given e.g. in Sanchez or Lemoine-Bouchard). In the same paragraph is a mention of Vigée Le Brun’s petition, which I published in 2017 with the essential Le Brun key allowing an analysis of all the signatories (why did so few women artists support her? – a question that only arises in the context of the full list).
I was disappointed to see Mlle Carraux de Rosemond’s name continuing to be misspelt throughout the book, despite my research on her biography (see her entry in the Dictionary of pastellists and this blog post), although the index entry (p. 357) indicates that my correction was known. And while it is correct to refer to Mlle Capet’s parents as “domestiques” at her birth (p. 55), my research has revealed that her father was soon after described as “homme d’affaires chez M. de Meximieu” [Jean-François Trollier de Fétan, seigneur de Messimieux (1731–1814), conseiller du roi en la cour des Monnaies de Lyon; a philanthropist and patron of the arts]. All of this is material to a discussion of the social status of women artists. Taking account of new research is what lifts surveys beyond the level of so much work that merely rehashes known information.
I noted dozens of other examples as I turned the pages, individually trivial, and mention just a few here to explain what I mean.
p. 25: “Miss Gardiner (fl. 1762–70)” (who John Hayes guessed might be a relation of Gainsborough, Susan Gardiner, when discussing an album now in the British Museum, while Wendy Wassyng Roworth called her Eliza in the 1997 Dictionary of Women Artists) was identified by Walpole as “the sister of Mrs Maccartney” [sic]. A letter from her in the Royal Academy archives gives her address as Upper Brook Street. She was in fact Mary Gardiner, sister of Henrietta, Mrs Francis Macartney, their brother Charles and nephew Luke, Viscount Mountjoy being significant patrons (with portraits by Cotes, Gardner, Reynolds) etc., and her niece Florinda was the girl shown dead in Maria Cosway’s picture (fig. 65).
p. 25: In some cases information is taken from secondary sources which is simply wrong: e.g., the death of Mary Grace, née Hodgkiss (Mrs Thomas Grace) is given as 1799/1800, as it appears in the Oxford DNB: but that is the date of the will of a homonym (Mary Harford, Mrs William Grace, mother of Clara Louisa Middleton). Just because it’s in the DNB… (In fact her maiden name seems to have been Hotchkis, not Hodgkiss; Lionel Cust, in his original DNB entry, thought she died in Homerton, Middlesex, c.1786, which would be consistent with rates books showing her in St Thomas’s Square until that year; the erroneous alteration was made in Marcia Pointon’s 2004 revision.)
p. 64: The painter Anne-Zoé Delaroche (1783–1850) married Guillaume Philippon Delamadeleine in 1818, aged 35: a bit more information than fl.1806-24.
p. 64: Françoise Windisch (b. Kilber, fl. 1801–8): Francisca Kilber married Johann Windisch in Höchst in 1776 and was active far earlier.
p. 38: Charlotte-Denise Surugue (1751–1820) was the daughter of the well-known engraver and academician Pierre-Louis Surugue. It appears (from the lengthy dispute with the Académie de Saint-Luc) that she and her sister specialised in hand-colouring engravings, in wash or gouache.
“Mme de Hauré (fl. 1776)” exhibited also in 1777. She was Marie-Jeanne Legrand (1750–1843), wife of the sculptor Jean Dehauré, pupil of Jean-Baptiste Lemoyne.
I won’t try your patience any further. Instead I will leave you with a final question, about a picture which is not mentioned at all in the book. It is the portrait of a young Black in fashionable clothing, now in the Cummer Museum in Florida, once supposed (without foundation) to be Mme du Barry’s servant Zamore, and signed and dated “Lemoine pinxit/a Paris mars 1785”. But which? Was it painted by Marie-Victoire Lemoine or by Jacques-Antoine-Marie Lemoine? (Incidentally it’s no. 56 in my catalogue of the latter’s work, published in the Gazette des Beaux-Arts in 1999; and it’s MVL 11 in Joseph Baillio’s catalogue of the former’s work, published in the same journal in 1996.) Space doesn’t permit a proper discussion of the reasons for our different conclusions: but the question that is relevant is this: does it matter whether the painting is by a man or a woman?
Postscript (13 August 2022)
I should perhaps take this opportunity to point out a commonly seen oversimiplification regarding the admission of women to the Académie royale in Paris: so often we read simply that their number was limited to four – or, as one recent book put it, there was “an official ruling that female members should never exceed four at any time.” It’s worth examining the exact wording of the resolution passed 28 September 1770:
L’Académie ayant considéré que, quoiqu’Elle se fasse un plaisir d’encourager le talent dans les femmes en en admettant quelques-unes dans son Corps, néanmoins ces admissions, étrangères en quelque façon à sa constitution, ne doivent pas être trop multipliées; Elle a arrêté qu’Elle n’en recevroit point au delà du nombre de quatre, si ce n’est cependant au cas où des talens extraordinairement distingués engageroient l’Académie à désirer, d’une voix unanime, de les couronner par une distinction particulière. L’Académie au reste ne prétend pas s’engager à remplir toujours le nombre de quatre, se réservant de ne le faire qu’autant qu’Elle s’y trouvera déterminée par des talens véritablement distingués.
I read this not so much as male malevolence as a shortage of really talented applicants (for reasons that are well rehearsed). But had another Rosalba Carriera presented herself, she would not necessarily have been rejected.
It has been my practice recently to turn my blog posts on art historical subjects into more formal essays on my website. Since they too are online, they can be updated and improved when any of you contact me with errors or new information. It’s then something of a chore to keep updates in sync. I also enjoy using Word and creating pdfs rather more than the horrid WordPress experience, particular after recent changes which make it even harder to control the layout and typography. So if you want to read what was to be my latest blogpost, about the collector Claude-Léger Sorbet (who was involved with Pigalle, Boucher, Cochin, Greuze – whose portrait, above, is in the musée des Beaux-Arts, Orléans), may I invite you simply to download the pdf from my site – here is the link: http://www.pastellists.com/Essays/Sorbet.pdf
The sad story of La tour’s cousin[1] has been told on numerous occasions, including by me, either erroneously or at least incompletely. For reasons that will be obvious, some parts of it will always remain unknowable, but a group of documents discovered in the last few weeks while researching a different matter merit revisiting the whole episode. As usual reference should be made to my chronological table, where transcriptions (but not facsimiles) of the documents may be found in chronological sequence together with full references. There is also a useful genealogy for La Tour, with this simplified version:
The story which appears in every account of La Tour’s life concerns his liaison with his cousin Anne Bougier, her pregnancy and the birth of her illegitimate child, for which as we know La Tour felt permanently guilty, and for which he made amends through his philanthropic donations many years later. The basic facts are found in the judicial interrogation of Anne, a document discovered by président Combier and published by him in La Petite Revue in 1874, and subsequently by Charles Desmaze in his Reliquaire de Maurice-Quentin de La Tour in 1874 (the original documents were presented to the musée at Saint-Quentin, but lost after being sent to Maubeuge during the First World War). This is Combier’s transcription:
Du novembre 1723. A comparu Anne Bougier, âgée de 22 ans, fille de Philippe Bougier, chantre en l’église métropolitaine de Sens, où il demeure à cause de son emploi, et d’Anne de La Tour, sa mère, avec laquelle elle demeurait en cette ville [Laon], depuis huit mois, et auparavant, demeurant l’une et l’autre, sa mère et elle en la ville de Saint-Quentin, n’ayant, non plus que sa mère, d’autre métier que celui de tricotter des bas.
A dit: qu’elle était née à La Fère, mais que sa famille était originaire de Laon. Feu Nicolas Bougier, Chantre en l’Église Collégiale de Laon, étoit son ayeul paternel, et feu Jean de La Tour, maître maçon à Laon, étoit son ayeul maternel.
A dit: qu’elle s’étoit bien comportée, n’avoit jamais eu d’habitudes criminelles avec aucun homme, ni garçon, à l’exception qu’elle s’est abandonnée trois fois au nommé Quentin de La Tour, garcon de dix-neuf ans, peintre de son métier, demeurant à Saint-Quentin, son cousin germain, et cela, dans le temps qu’elle demeuroit avec sa mere à Saint-Quentin.
Interrogée si c’est des œuvres dudit de La Tour, son cousin, qu’elle est devenue enceinte de l’enfant mort, dont elle est accouchée le 15 août 1723, après avoir célé sa grossesse, a dit que oui, qu’elle s’est crue hydropique, parce qu’après avoir eu ses habitudes avec le dit de La Tour, elle a eu ses purgations ordinaires huit jours après et ne les a plus vues depuis.
Anne Bougier, ne sachant signer, est déclarée atteinte et convaincue d’avoir tenu sa grossesse célée jusqu’au jour de ses couches et, pour ce fait, condamnée a être admonestée en la chambre du Conseil à ne plus récidiver, et en 3 livres d’amende, applicables aux pauvres de l’Hôpital de Laon.
Lapauze (1919) went so far as to state that she was “faite prisonnière” by La Tour, and that evidently was the view of the tribunal reflected in her punishment (concealment of pregnancy was regarded as infanticide under an edict of 1566). According to her baptismal record (8 mars 1700), only located in 2019, she was in fact 23½, four and a half years older than La Tour: a difference in age making this defence somewhat less plausible than if he had been older.
No doubt the pathos of the story inspired genealogists to try to complete the picture, not always helpfully. The normally reliable Maurice Tourneux this time was responsible for repeating information he received from Jules Hachet in 1904, subsequently widely repeated by modern authors – including by Christine Debrie in 1991 (and of course still polluting genealogy websites). According to the story the unfortunate girl did marry, soon after the affair with her cousin, and settled down with her husband, a workman called Bécasse, in the parish of Saint-Thomas in Saint-Quentin where she died in 1740. I compounded this by finding an earlier register entry for the baptism of a child from this legitimate marriage, in 1728. But examining these entries carefully, they don’t refer to a Marie-Anne Bougier at all, but to a Marie-Anne Bruge or Bruche: the writing in each case is quite clear. It’s neither a likely phonetic mistranscription nor a likely pseudonym if she wanted to disguise her past; nor do the witnesses seem to have any connection with the pastellist’s family. And the age given at her death was 45, so that she would have been born in 1695.
In 2016 I made one further discovery, which I find almost as disconcerting: as we know she was the daughter of the pastellist’s aunt, Marie-Anne de La Tour, who married a Philippe Bougier, a fellow chantre in the church. The marriage took place in Laon in 1695 (17 May) when Philippe, a widower, was 26 years old (which was one of the reasons I continued to believe Tourneux’s identification). But I’ve since located Marie-Anne de La Tour’s baptismal entry:
She married Bougier when she was barely twelve years old. This was no dynastic match in which contracts were entered between children to be consummated when they reached adulthood. There is likely to have been a pressing reason, but whether it was an unrecorded sibling of Anne Bougier the registers do not vouchsafe.
* * *
Ever since the publication of La Tour’s wills, there has been something of a puzzle concerning the beneficiaries he describes as his “cousins”, almost all of whom I identified in 2016. But one of beneficiaries named in his later (1784) will that remained stubbornly unexplained was a “Mme La veuve Grand Sir, a La Ferre en Picardie” (La Fère):
Despite spending a vast amount of time in numerous archives and websites trying to unravel this in 2016, I stumbled on the answer only in June 2022 – in the parish registers of Saint-Montain, La Fère, when I was researching something quite different. This was evidence that there was indeed a Mme Grand Sire, or Grandsir, in La Fère, of an age that meant she might well have been a widow still there in 1784. Her name was “Barbe-Antoine Dio–” when, on 23 December 1750, she gave birth to a boy called Jean after his father, also Jean Grandsir, a tissserand in La Fère:
Le vingt trois a eté baptisé par moi chanoine Curé Doyen soussigné jean fils de jean grandsir tisserand en cette ville et de barbe antoine dio–– son epouse le parein jean du Notion la mareine francoise cheval, ledit baptisé né le jour meme
signé: De Nelle
It was evident that the curate had not been able to get the mother’s name correctly, but it was sufficient to make me return to the search for more details. This yielded the entry for the marriage of Barbe-Antoinette and her husband, Jean Grand Sire, tisserand, the previous year (1749), in Laon, Saint-Jean-au-Bourg:
Le vingt Janvier mil sept cent quarante neuf aprés avoir publie les trois bancs de mariage en deux Dimanches et vue fête entre Jean Grand Sire <homme veuf> fils de Jean Grand Sire Maitre Tisserand, et de Margte Guilbert demt a Aubegast, diocese de Roüen, d’une Part <age de 38 ans> et de Barbe Antoinette Guiot fille d’Anne Bougier demt a Laon de Cette Paroisse d’autre part <agee de 24 ans> Sans qu’il soit venû a ma connoissance aucun empechemt qui puit retarder la Celebration dudit mariage Je Soussigné Jean Antoine Huët prétre licentié en Theologie de la faculté de Paris, Curé de la Paroisse de St Jean au Bourg de la Ville de Laon, ay recûs de Jean Grand Sire et de Barbe Antoinette Guiot les promesses et Consentemens de Mariage et l’ay Celebré en l’Eglise de laditte Paroisse avec les Ceremonies accoutumés en presence de Jean Charles Marteau clerc laïc de la paroisse de St Michel, d’Antoine Larmois Clerc laïc de laditte paroisse de St Jean au Bourg de Nicolas Taïtart Me bonnetier et de Felix Bon bion Vigneront, dems tous en cette Ville soussigné avec L’Epoux et l’Epouse qui onts signés aussi le Jour et an Susdits
signé: jean grandsire barbe antoinette guiot Marteau
tetard felix bion Larmois
huet curé
From which we can see that Mme veuve Grand Sire was in fact Anne Bougier’s second child, born almost certainly in 1724, the year after the stillbirth of La Tour’s child that caused the trial discussed above. But the format of the entry is far from standard, and the acte leaves open many questions. Minor children (any unmarried person under 25) could only marry with their parents’ consent (and normally their presence at the wedding), so it is extraordinary that Barbe-Antoinette Guiot, aged 24, married without any father being named, nor it seems with her mother present (or identified as deceased). Evidently she was illegitimate, the father unknown. Was her mother dead by the date of the marriage? Could “Guiot” come from the name of a biological father, a stepfather or a protector? (It seems possible that this was Gérard Guiot or Diot (both names appear in the records), born 1680, a maître boulanger or patissier in Laon; on 23 November 1705 at Sainte-Benoîte, Laon he married a Barbe-Nicolle l’Eully.)
Further research in La Fère registers produced another baffling document: Barbe-Antoinette’s acte de décès, in 1792, claiming to be aged 83 which would make her far too old as well as contradicting the 1750 acte de marriage.
Sur la déclaration a nous faite par la citoyenne Marie Auteffe, demeurant à l’hopital des pauvres de cette ville, en qualité de surveillante desdits pauvres, agée de cinquante trois ans; que la nommée Antoinette Diot, veuve de Jean Grand Sire, cavalier de maréchaussée du Soissonnais à la résidence de La Fère, agée de quatre vingt trois ans natif de Laon, chef-lieu du département était décédée du jour d’hier, à cinq heures et demie du soir audit hôpital…
Was this “veuve de Jean Grand Sire” a different woman? I don’t think it can be as the name Diot had already appeared, she is described as from Laon and Jean Grand Sire is not a common name in La Fère.
I then uncovered yet another piece in the jigsaw, this time in the parish register of Saint-Rémy, Dieppe (surprisingly distant from the other towns we are concerned with: Saint-Quentin and Laon are within a 25 km radius of La Fère, while Dieppe is 200 km away), two years before Barbe-Antoinette and Jean Grand Sire’s marriage of 1749, once again filled with inaccuracies, whether erroneous or deliberately intended at concealment, but with sufficient contiguity to the truth to tell its own story:
Ce jeudy vingt-huit de decembre fut baptisé par monsieur Feburier vicaire Jean Charles fils illegitime né de ce jour de barbe anthoinette deLatour originaire de Lion en Lionnois fille de feu Jean de la tour et de marie anne bouzier de cette paroisse provenu des œuvres de jean grand sire aubergiste Suivant la declaration passée devant monsier charles adrien de quiefdeville bailly juge civil criminel et de police du trois d’octobre dernier portée au mandement et datte de ce jour signe de quiefdeville avec paraphe legris avec paraphe et scellé, nommé par charles gachet soldat invalide de la compagnie de monsieur beranger en garnison au château de cette ville de cette paroisse, et Anne bougier veufve de jean delatour fileuse de la paroisse de Saint michel de lon en lannois Le parein a signé Ledit jean grandsire absent La marreine a fait sa marque en declarant ne sçavoir ecrire
signé: Charles Gachet La marque d’anne bougier + qui a dit ne scavoir ecrire
feburier vicaire de St Remy
So Barbe-Antoinette herself had an illegitimate child before marriage, just as her mother had done. But the document sheds important new light: firstly that Anne Bougier was still alive, and present (and still unable to write): from tricoteuse de bas she had become a fileuse. Moreover while the infant’s father was the Jean Grand Sire who would later marry the mother, our attention is engaged by the name Anne gives to Barbe-Antoinette’s father: Jean de La Tour, claiming to be his widow, thus explaining his absence. Any other claim would easily have been exposed, Maurice-Quentin de La Tour being by then a name quite likely to be recognised by a vicar. “Jean de La Tour” is a name so common as to be untraceable, particularly before Google, but the only Jean in the pastellist’s family was his grandfather, far too old to have fathered Barbe-Antoinette.
It seemed worth trying to investigate Jean Grand Sire’s background. Evidently he had a portfolio career: an aubergiste (1747), maître tisserand (1749, 1750), and later cavalier de maréchaussée du Soissonnais à la résidence de La Fère. According to the 1749 acte de mariage, he was the son of another Jean Grand Sire, another maître tisserand from Auppegard near Rouen, and his wife, Marguerite Guilbert, already a widower and born c.1710. Those people exist: Jean Grandsir (1685–1767), who married Marguerite Guilbert ( –1758) in Colmesnil-Manneville (10 km south of Dieppe) on 25 November 1710 and died in Auppegard, a further 1 km south of Colmesnil (within the same parish). (I also came across Charles Gachet’s signature again as a witness in the Colmesnil parish register.) They had several children but none called Jean is recorded. It is quite possible that he was born before his parents’ marriage, which didn’t take place until near the end of the year, accounting for the absence of a baptismal entry. Confusingly a Nicolas Grandsire and Marie-Suzanne Guilbert also had children baptised in the same parish around the same time.[2]
While I was proof-reading this article I decided I’d better have another trawl through the Laon parish registers, just in case I, together with everyone from the président Combier, Maurice Tourneux, Charles Desmaze and everyone else had missed something. And we had. Here, almost exactly where you would expect it, is the acte de baptême of Marie-Barbe-Antoinette Guiot, in the parish register of Saint-Michel, Laon, dated 4 December 1725:
Le quatrieme jour du mois de decembre mil sept cent vingt cinq est nee et a etee baptisee marie barbe antoinette fille d’Anne bougier femme non mariée qui a declaré qu’Antoine guiot cordonnier en vie etoit le père dudit enfant elle a eü pour parein valentin fourfaux et pour mareine anne therese damour qui ont signet ou marquet avec moi le present acte les jour mois et an que dessus
+ marque de la mareine Valentin Fourfaut Agnet
So there was a Monsieur Guiot – a shoemaker, possibly dead (although Laon parish registers do not record such a death in the previous nine months) and possibly married (a Pierre-Antoine Guiot was married there in 1722; he signed Diot while his father signed Guiot: evidently the spelling caused his own family the same problems Barbe-Antoinette would later show).
One thing is clear. Anne Bougier’s transgression with La Tour was not an isolated incident. But why should she have (approximately) named La Tour as her second child’s father in 1747? Could it be that he was in fact the father of Barbe-Antoinette? Did she think the name a grander one for her daughter to bear? Or did she harbour some resentment at his conduct? I leave you to decide whether it affects your views of the artist’s moral character – and whether that has any relevance to his art.
NOTES
[1] This essay first appeared on 26 June 2022 as an update of Jeffares 2016j, incorporating material from that and substantially extending it with a discussion of Barbe-Antoinette. It may be cited as Neil Jeffares, “La Tour’s cousin Anne Bougier”, Pastels & pastellists, http://www.pastellists.com/Essays/LaTour_Bougier.pdf and is referred to within the Dictionary as Jeffares 2022c.
[2] Two homonyms lead to false trails: a Jean Grandsire had been born to Nicolas and his wife in 1714, but on 27 Nov 1736, still in Colmesnil, he married a Marguerite Sannier, was able to write – in a hand that does not match that on the 1749 acte de mariage, and so cannot be our Jean Grandsire but may have been a cousin. Another red hering is the Jean Grand Sire who married, in Dieppe, Saint-Rémy, on 18.i.1738, a Marie-Marguerite Baron; he was then described as a “pignère de profession” (a carder), aged 21, the son of Jacques Grand Sire and Hélène Le Coq, unable to write. The following year, on 21.v.1739 in the same parish he married Marie-Marguerite Maugendre, a dentellière aged 25, as Jean-Claude Grandsire. He was dead by 1749 when his second wife remarried.
Neil Jeffares 