Skip to content

The evolution of taste in pastels

18 October 2017

To accompany the new exhibition of nineteenth century pastels currently at the Petit Palais, L’Art du pastel de Degas à Redon, a two-day seminar Le Pastel: regards croisés sur une technique singulière was held in Paris on Monday and Tuesday. I gave two talks yesterday morning, of which the first is in this post. The second will appear later. My theme concerned two different aspects of pastel exhibitions. You can easily find full references, bibliographies etc. for the talk in the online Dictionary of pastellists, and in particular in the Prolegomena.


Let me first of all say a few words about the scope of my talk. I am discussing only pastels made before 1800. I consider nineteenth century pastels to be completely different objects which just happen to use a similar medium. Degas and his contemporaries did not revive the dix-huitième: they may have been inspired by it, to some degree, but they were led to something far removed visually and aesthetically.

But the history of connoisseurial appreciation of the dix-huitième pastel does of course go back before 1908, and we must start with that general revival of interest in 18th century art which is so often associated with the Goncourts.


Our understanding of this has advanced considerably thanks to the excellent La Caze exhibition 10 years ago, and the subsequent colloquium at The Wallace Collection in London entitled Delicious Decadence. I’m going to assume that you are familiar with both publications. But it may surprise you to learn that the word “pastel” hardly occurs at all in the proceedings: only one pastel is actually mentioned, and that in passing.

Carole Blumenfeld’s excellent article in the La Caze catalogue discussed a number of the pioneers in this rediscovery: Daniel Saint, His de La Salle, baron d’Ivry, baron de Beurnonville, Laurent Laperlier – although Henry de Chennevières subordinated their roles to just three figures: François Marcille, Louis La Caze and Hippolyte Walferdin. To these Blumenfeld judiciously added the marquis de Cypierre and Paul Barroilhet. But how far did their interest extend to pastel?


If we look at the collections they and others formed – and exclude groups of portraits in museums or still belonging to descendants of the sitters, as well as the stock of dealers – surprisingly few people acquired pastels systematically. My figures indicate only about 120 collectors owning more than four 18th century pastels, and of these fewer than 30 owned more than a dozen. Topping the list was indeed François Marcille, with more than 50 pastels, Camille Groult a close second. Among Blumenfeld’s figures, the only others with more than half a dozen pastels were Laperlier and Cypierre, of which Laperlier’s collection was more important. The other great collections – Doucet stands out – were formed a little later, when the revival had taken hold. But in almost all these cases, the substantial pastel holdings were matched – or rather trumped – by prodigious holdings of oil paintings or drawings. Marcille’s 55 pastels for example were part of a collection numbering at least 2700 works.


I don’t have the time to offer biographies of each of these figures. What I can say is that they had surprisingly little in common: there were painters and financiers, nobles and nouveaux. Some were prodigiously wealthy, some were not. Some were politically ultra, others progressive. Curieux indeed.

There is a further aspect to this: many of these nineteenth collections included pastels only by certain big names: La Tour, Perronneau, Chardin, Greuze, Boucher, Prud’hon, Fragonard. To these were added Rosalba, and by the turn of the century the English artist John Russell, invariably misspelt. Apart from those two and Perronneau, the others were all the subjects of the Goncourts’ essays in L’Art du xviiie siecle, which started to appear in 1859. And we have to ask how deep was their understanding of pastel – indeed how many of those late 19th century records of pastels by La Tour, listed but not confirmed in B&W and now disappeared without trace – were actually correctly attributed? This was an era when ludicrous misattributions and misidentifications were endemic.


In 1863, the Goncourt brothers visited princesse Mathilde and were shown into her painting studio, which was “encombré de ces choses qui ne sont des objets d’art que pour les femmes, un faux pastel de Boucher, de faux pastels de Chardin.” Just a year later, this pastel entered the Goncourt apartment in the rue Saint-Georges:


They had seen it in a sale where, like all other competent pastels, it was catalogued as by La Tour, a description they recognised as false. The brothers were excited to find it on their return from a fencing lesson:

Un commissionnaire nous apporte des Commissaires-Priseurs un magnifique pastel de Perroneau, pour lequel nous avions donné commission, dimanche, à une vente de tableaux de l’École française. Nous nous habillons, mettons des cravates blanches, allons dîner chez la Princesse, revenons et fumons une pipe, en adoration devant notre Perronneau, posé sur la table de notre chambre.

The magic took its effect: the entry for New Year’s Eve is:

En regardant le Perroneau et nos tapisseries de Beauvais, je songe que le xviiie siècle a eu, dans son ameublement d’art, le velouté.

But of course the pastel they had bought was not by Perronneau either, despite the label they affixed to the frame. They paid Fr 330 francs in 1863; at their sale 34 years later it fetched ten times as much, Fr 3000 – still only about €15,000 after inflation in today’s money. It is in fact by Vigée Le Brun.


Remember too Reynaldo Hahn’s account of the Goncourt view on Perronneau’s superiority to La Tour and Chardin – the “other two pastellists”; and that travel to find out more about other schools was quite unnecessary. Was this based on deep understanding – or rather social bluster, claiming an exquisite sensibility denied to lesser mortals? This was 1895: although the Louvre had recently acquired its third Perronneau, Hahn and Proust had only seen La Tour and Chardin on their visit the day before this dinner. Proust wrote his unpublished essay about Chardin at this time, but he mentions Perronneau in his novel only after visiting Cent pastels in 1908.


It’s a sobering thought, but perhaps one that should not surprise us. Because quite simply the opportunity to see pastels in public collections was very limited in the mid-19th century. The collection at Saint-Quentin was barely accessible until it was placed in the Palais de Fervacques in 1856, and only moved to the present musée Antoine-Lécuyer in 1886.

The collection at the Louvre was of course the most important, and certainly well copied by the end of the 19th century. We can begin to glimpse the evolution of taste by looking at the holdings in the Louvre from its inception in 1793 when it took over the royal collections and those of the Académie royale. Excluding sheets with touches of pastel, the Louvre now owns about 200 pre-1800 pastels, representing the work of approximately 70 pastellists of whom these are the top dozen:


These names probably won’t surprise you, although the exclusions are remarkable. There was no Liotard pastel until 1982, and there is still only one. Apart from a handful of Russells, the neglect of non-French pastellists is systematic: nothing by Copley, Cotes, Gainsborough, Gardner, Hamilton, Hoare, Mengs, Rotari, Schröder or Tiepolo.

But we should also look at the date of acquisition of these holdings.


The peak in the 1820s is probably a glitch in the inventories, and should probably be added to the ancien régime endowment representing one-third of today’s holdings. That in the 1940s includes wartime recuperation. But the peaks in the 1860s, the 1910s and the last 15 years do represent waves of real interest. Of the 27 La Tours, a dozen were there at the start, another ten were added around 1910, while the 3 additions since 1950 have all taken place in the last 12 years.


Over the years numerous pastels were offered to the Louvre: there have been some astonishing refusals – among them the président de Rieux now in the Getty, as well as the abbé Nollet in Munich. Even the full length Mme de Pompadour was initially refused before the Louvre changed its mind. Mme Roslin’s self-portrait was offered but rejected in 1847. As we shall see that was a mistake with sad repercussions.

One might expect that museum acquisitions worldwide would provide a valuable chart of leading taste across all schools. Sadly that seems not to be the case. Only a handful of museums have both the funds and the skills to set trends in taste. For others the pastels they acquire have been through legacies or random acts that reflect specific interests (for example when the sitter in a portrait has a connection with the museum) rather than a planned programme of acquisition. I have records of some 6500 pastels in public collections worldwide, so to sift these I made an arbitrary decision to look at acquisitions only in the last 30 years. I can’t say I pursued this with 100% accuracy, but I found details of some 300 pastels acquired in that period by 104 institutions. But when you exclude those acquiring only a single or couple of works, the number reduces to just 20 institutions worldwide:


Even here, I’m including the V&A which received two wonderful Nattier pastels by legacy 30 years ago which have never, and probably never will, come out of storage: they are an exception to the general rule that British institutions dislike all pastels, and hate French in particular. But the names topping the list will surprise no one here.

Equally there are no surprises when we analyse the acquisitions by the names of the artist, where perennial favourites dominate:


Among the 85 pastellists, only two dozen were represented by 4 or more works, and even these numbers are distorted by special interests. Thus Guildford House and Geneva respectively collect Russell and Liotard. Perhaps the most curious phenomenon was the interest shown by a number of minor US galleries in acquiring single heads by Luti from a group that appeared at auction some years ago. Many of these acquisitions were driven by skilful dealers.

Art market

Their influence can be followed in the saleroom. You can find more detail on all this on my website, and I don’t want to get bogged down in dry statistics or arguments about inflation. But here is a summary of the pastels which sold for more than $100,000: there are 89 pastels by 29 artists.


Prices for pastels collapsed at the end of the eighteenth century and only really revived with the second wave of interest towards the end of the nineteenth century, reaching a zenith in the first quarter of the twentieth century.


We can see this clearly in individual cases, such as that of a Perronneau pastel (left) which belonged to Laurent Laperlier, but fetched a mere 48 francs (about €200 in today’s money) in his 1867 sale. Auction prices are made by two competing bids, and a sole enthusiast does not represent a market revival. The purchaser was the other enthusiast, the baron de Beurnonville. In 1925 it sold for a more respectable 17,000 francs. Marius Paulme bought, and at his sale four years later it went for 70,000 francs (€175,000) – about 1000 times more than at the Laperlier sale. Today it would be worth rather less.

The Perronneau known as the comte de Bastard in the Louvre was sold in 1881 for 5000 francs (about €23,000). Seventeen years later it was still only worth €40,000 in today’s money. But at the Jacques Doucet sale in 1912, it fetched 128,000 francs, equivalent to nearly €600,000 today.

The Jacques Doucet sale in 1912 marked the high point in French pastels, with La Tour’s Duval de l’Épinoy (right) reaching about €3 million today, double the estimate. Writing in the Burlington magazine, Robert Dell, its first editor, revealed typically British fury: “Is it in accordance with common sense that a masterpiece by Fragonard [le songe du mendiant] should fetch 137,500 francs, and a masterpiece by Latour, who can hardly be counted the equal of Fragonard, 660,000? The truth is that prices have no sort of relation to artistic value.” But they do tell us about taste.

This revolution in saleroom prices, which lagged some way behind the development of individual enthusiasms, occurred simultaneously in London and Paris, and applied to English as well as to French pastellists. At the Angerstein sale at Christie’s in 1896, the 1000 guineas (€130,000 today) achieved by a Lawrence (one which had taken him all of three weeks to complete) was the highest price ever paid at auction for a pastel. It was soon exceeded by Russell’s Persian Sibyl, for which Charles Wertheimer paid 1100 gns (Christie’s, 1899; about €150,000 today). The 1908 sale of the Gardner portrait of Lady Fawkener for 1250 gns (again about €150,000 today) brought the artist out of obscurity, and created a self-fuelling dealers’ bonanza – although in fact a Gardner had been sold four years previously, at the Townshend Heirlooms sale, when it was miscatalogued as by Reynolds.


There it fetched 35 times as much as a Rosalba, while the unattributed pastel by Katherine Read still went for four times the price of the Rosalba.

These relativities are evidently unstable, and these prices were not sustained. It was not until 1993 that a British pastel reached a level equivalent to six figures today. Since then Hugh Douglas Hamilton and chalk drawings by Gainsborough and Wright of Derby have become popular.


Perhaps the most surprising performer has been Liotard. Although known for high prices when he was working, his masterpiece, Le Petit Déjeuner des Mlle Lavergne, sold in 1801 for £89 (about €8000); in 1835 for £31 (€4000); in 1916 for £1260 (€130,000); and in 1918 for £1450 (€100,000). No other published price for a Liotard reached that level until the 1986 purchase by the Museum of Fine Arts in Houston for a reported SwFr 2 million (€2,400,000) – but that was a response to the Getty’s purchase of the little girl, Frederica van Reede, for an undisclosed sum a few years before. Thereafter some 15 Liotards have sold for over £100,000, dominating the tables by number and value.

Liotard’s market value coincides with the burgeoning literature devoted to him. But it cannot be said that academia has played a significant role in the taste for or even in the study of pastels. There is almost no appetite for object led research. So we need to turn to public exhibitions,



The salons of the Académie royale have of course been the subject of much academic research – although those of the Académie de Saint-Luc, far more important for pastel portraits, have been largely overlooked. But these were shows of contemporary art. Pastels were mixed with paintings, arranged by size and format.

During the 19th century, the neglect of 18th century pastels in loan exhibitions reflects that in the saleroom.


If we look at loan exhibitions including more than ten 18th century pastels, worldwide, I can find only 15; of these, the six in Paris all took place after 1874. These too were heavily dominated by pastels supposed to be by La Tour. There may have been one or two by Coypel, Rosalba or Perronneau, but no pastel by Vivien, Valade, either Vigée, Labille-Guiard, Lenoir, Loir, Lundberg, Boze, Ducreux, Glain, Nattier: these were all unknown as pastellists.


The Exposition de Cent pastels represents the start of attention on the 18th century pastel in its own terms. It changed a great deal:

  • It made collecting pastel socially acceptable
  • It introduced the medium to a broader audience
  • It introduced them to minor pastellists such as Jean Chevalier, Duplessis, Frey, Guérin, Hall, Hoin, Labille-Guiard, Mérelle
  • It even made English pastellists such as Cotes and Gardner acceptable.

Standards of scholarship however left much to be desired, as anyone can see by comparing the livret with the commemorative catalogue, where at least two dozen new “identifications” were supplied by Roger-Milès’s fevered imagination. But perfectly good works (such as La Tour’s princesse de Rohan,


now in Stockholm) were rejected as anonymes, in favour of copies and misattributions. As a quick tally, of the 118 pastels in the livret, only 102 made it into the final catalogue, not all of which are known today; I reckon that some 25 were given imaginary identities, perhaps ten were given to completely incorrect artists (Vigée Le Brun to Carmontelle or Perronneau; Valade to La Tour; Gardner to Reynolds etc.), while a dozen (mostly given to La Tour) were simply copies. The point was made by Maurice Tourneux at the time, and more recently by Xavier Salmon in the context of the 2004 La Tour exhibition. I make it again in the broader context of “not every pastellist is La Tour”.

But Cent pastels was a start. It was rapidly followed by the even larger Exposition des pastellistes anglais du xviiie siècle – but this was a dealers’ show, riddled with error and diluted with watercolour. The First World War intervened, but by 1920 there was a return of interest led initially by dealers’ shows but culminating in a re-run of Cent pastels, but with sounder scholarship: the 1927 Exposition de pastels français des xviie et xviiie siècles.


Again a livret and a full souvenir catalogue, and this time as well as including further unknowns, such as Allais, Berjon, Davesne and Frédou, an attempt was made to cover the earlier period, with pastels by Nanteuil, Vaillant and Vivien. The English interlopers – indeed all foreign pastellists – were excluded. There remained some errors: among nearly 150 pastels, about a dozen were wrongly identified and a similar number misattributed. Despite the inclusion of many smaller names, it reinforced the perception of La Tour and Perronneau as the only ones who matter. In general it established a strong level of interest in 18th century French pastel, and led to the inclusion of pastels in a series of loan exhibitions in Paris, Amsterdam, London, Vienna and Venice in the 1930s.


Paul Ratouis de Limay finally published his survey of French 18th century pastellists in 1946, but it was largely based on work done before the war (he was 65 when it appeared). And while there were similar, but even less accurate, summaries of German (Brieger, 1913) and English pastellists (Sée, deeply flawed) there remained no comprehensive study of the field until far later.


The next war cast a pall over the dix-huitième which lasted through the 1980s. You can see too what happened in terms of cross-border movement: although a couple of dealer-led shows in Paris and London like the Pastellistes anglais distorted the statistics for the 1910s, the transport of pastels for international loan exhibitions was very rare until very recently. This is a point I shall come back to.

Of course it’s not correct to say that pastel shows ended in 1939. Pastels continued to appear in more general exhibitions. The huge Marie-Antoinette show in Versailles in 1955, with over 1000 exhibits, included some 30 pastels. (In contrast the 2008 Marie-Antoinette exhibition included only 6 pastels.) A similar number and ratio for the Royal Academy exhibition of British portraits in London in 1956. But pastels in larger shows took a lower and lower profile. The Royal Academy’s winter exhibition of 1968, France in the eighteenth century, had 1035 exhibits, but only 5 were genuine pastels. Portraits publics, portraits privés at the Grand Palais in 2006 and the Royal Academy the following year included no pastel.

Apart from a few commercial shows pastel exhibitions were rare in the post-war period. Special shows of the Louvre and Saint-Quentin collections were held in the 1950s; the next exhibition with more than a dozen pastels was that of the Carnavalet’s permanent collection in 1984. After that you had to wait 13 years for the Versailles exhibition, again of its own holdings. In 2001 the Hermitage did the same.


By far the largest was the Warsaw exhibition of 2015 with some 250 of its own pastels from all periods.


2004 saw the tercentenary celebrations for La Tour with 50 pastels from public and private collections travelling to Versailles and exhibitions of permanent collections at Chantilly, the Louvre etc. with a joint catalogue: the simultaneous hub idea which never really caught on. But the event gave a huge fillip to pastel exhibitions.


Recent years have seen a number of monographic exhibitions: Rosalba, Tiepolo, Liotard, Perronneau.


The Vigée Le Brun show had some 32 pastels in Paris, although only 5 made it to the US leg. But a feature of this and the Liotard and Perronneau shows which may be termed “availability bias” is that for these artists, the relative strength of their œuvre in pastel compared with their oil paintings may have been distorted by the difficulty of borrowing their best work in the more fragile medium. Some people left the Liotard exhibition thinking he was a far better miniaturist, painter and draughtsman than pastellist.


The most important recent shows devoted exclusively to the eighteenth century pastel by multiple artists was held by the Met in New York in 2011: 44 pastels of its own and local collections, mainly focused on big names, but mixing pastels from all schools.

But the need has remained for loan exhibitions so as to extend knowledge beyond domestic boundaries and the inner circle of the big names.

This survey prompts questions:

  • Should French pastels be mixed with British, Italian, German and other schools?
  • Should 18th century works hang with later ones?
  • Should pastels hang among oil paintings, drawings or miniatures?
  • Should they be hung sparsely in single rows, or in dense multi-level displays?
  • How should they be lit (ambient or directional), and against what colour walls? Should there be shows with different levels of illumination: one night a month reserved for pensioners, or fifteen minutes each hour at general low levels?

Let me try to draw some conclusions.

  • Taste is influenced but not formed by individual enthusiasts. Whether they are ahead of their time or remain disregarded is, like treason, a question of dates.
  • Museum purchases, saleroom prices and social acceptability are all driven to a far greater extent by dealers and perhaps increasingly by auction houses.
  • Taste has been and remains strongly name-driven. A Gardner or even a Liotard is nowhere until someone pays a spectacular price.
  • European collections largely remain national. Only from America is Europe viewed as a single country.
  • Neglect and ignorance persist, and exhibitions are crucial in breaking this down.
  • But fragility remains the big barrier to those shows…my next talk.

From → Art history

  1. Reblogged this on Silicon Valley Types and commented:
    A scholarly taste for pastels will find this a marvelous blog. Others may pass–but do enjoy all the pastel portraits displayed before you go!

Trackbacks & Pingbacks

  1. Identifying Russell’s petite fille aux cerises | Neil Jeffares

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: