Impressionism

To the Editors:

In view of Jack Flam’s doubts about social art history, his review of my Impressionism: Art, Leisure, and Parisian Society [NYR, September 28] is very generous. I can only be pleased that he grants to the book many positive features. He nonetheless unwittingly misleads your readers on several counts. He complains that I omit Cézanne, Sisley, and Pissarro, and that I don’t deal with portraits, “pure” landscape, and still lives, as though these were unthinking shortcomings on my part. In my preface I write that, in view of my focus on leisure, “portraiture, landscape, and still life do not figure here in their own right.” Compare Professor Flam: “Portraits, pure landscape, and still life painting are almost entirely overlooked.” My book makes no claim to deal with Impressionism tout court. Instead I study only one facet of the movement: the painters’ devotion to Parisian leisure and entertainment (and extensions outward to suburbs and seaside). Cézanne, Sisley, and Pissarro weren’t involved with Parisian leisure.

Flam says that my book is not “the worthy successor to Rewald’s History of Impressionism.” I had no intention of supplanting John Rewald’s magisterial biography of the movement. Precisely because it is available I felt free to write topically rather than chronologically, since my readers can rely on Rewald. I realize that Flam wishes I had attempted an all-encompassing study of Impressionism, but I would like your readers to know that the focus of my book is quite intentional. A minor note: Flam complains that the sizes of the pictures are omitted in the captions, but neglects to say that the interested reader can find them in the list of illustrations.

Our differences are those that usually crop up when an art historian regards social history as a kind of contamination, and therefore concludes that discussions of “social background” are separate from art. Flam’s phrase reveals the nub of our disagreement. Social history is in intimate and complex dialogue with art; it is not the “background” that Flam would have us believe. He is unwilling to see joined together what he truly believes on principle to be sundered. He wants to cling to the idea that art is elevated above social history to a separate realm of “poetical experience.” Poetical it is, separate it is not.

Allow me to give one final example of our differences and simultaneously right the record. Flam accuses me of “cavalier” use of sources by not emphasizing that Rivière’s discussion of the models in Renoir’s Dance at the Moulin de la Galette was published in 1921. (Of course I footnote the source, and Rivière’s earlier account of the picture.) I use the 1921 account precisely because then, long after the fact, Rivière lets us know that the men portrayed in the picture are, like Renoir, visitors to Montmartre, artists and writers who use the local women as models, although in his account of 1877 Rivière lent himself to the idea that the picture shows an ordinary Sunday ball at Debray’s. The account of 1921 unwittingly undermines the myth in which Renoir and Rivière so earnestly believed in 1877, the “general and typical nature of the painting” that Professor Flam wants to preserve.

Robert L. Herbert
Department of the History of Art
Yale University
New Haven, Connecticut

Jack Flam replies:

How I have misled readers by characterizing the scope of Robert Herbert’s book almost exactly as he himself did is beyond me. Just because Herbert was aware of the limitations inherent in his view of Impressionism doesn’t mean that they are beyond criticism. Moreover, his somewhat arbitrary focus on only those aspects of Impressionism that serve to make his points presents its own problems. For it suggests that perhaps not all Impressionist art is equally subject to analytical methods that take into account the social setting. And this I would take strong issue with, since, contrary to Herbert’s assertions, I firmly believe that all art can profitably be studied in relation to its social setting.
Regarding the comparison to Rewald, I don’t think that Herbert is blameless here, since the title of his book implies a broader scope than it has. If a book’s title does not accurately reflect its subject matter, the author can of course change that title instead of resorting to elaborate and not completely convincing disclaimers in his preface. And any author who is really concerned about misleading readers would do well to pay close attention to how his book is described on its dust jacket. On Herbert’s book, the dust jacket copy begins with the following quote: “This full-scale revision of Impressionism immediately supercedes all other studies in the field.” Such statements, of course, do not exist in a vacuum and I am not the only person who has been moved to compare Herbert’s book with Rewald’s.

Herbert’s remark about the captions indicates that he seems unable to refrain from rationalizing virtually any criticism of his book, however small. Petty as this defensiveness may be, however, its implications are not so minor. My basic assumption, evidently different from his, is that all readers are “interested” in what they are reading about. And since in this case they are reading about pictures, they can quite reasonably be expected to want the sizes of the pictures to be easily accessible in the captions. Awareness of a picture’s size is crucial to understanding it, and Herbert’s negligence in this matter reinforces one’s suspicions about his apparent disregard for the specificity of individual paintings as works of art.

I believe that “cavalier” is a fair and not unkind characterization of Herbert’s use of Rivière’s texts about Renoir’s Dance at the Moulin de la Galette. It is much kinder, in any case, than the way one would have to characterize Herbert’s defense of his use of those sources. For nowhere in Herbert’s description of the picture does he tell us whether his account is based on Rivière’s 1877 article or on Rivière’s 1921 book—and his footnotes are, to put it mildly, rather vague on this matter. In fact, Herbert never even points out that Rivière wrote two substantially different texts about the painting. Herbert’s detailed discussion of the painting is preceded by a footnote which refers to Rivière’s 1921 book (as “the best account of the site, its people, and Renoir’s painting”), and is followed by a footnote referring to some misleadingly used quotes from Rivière’s 1877 article. (Rivière’s remarks are linked to the models Jeanne and Estelle by name, even though Rivière gave no names in his 1877 article.) Herbert not only doesn’t differentiate between Rivière’s two descriptions of the painting but combines them to buttress his own argument.

Specifically, after Herbert’s detailed description of the people in the painting (based largely on Rivière’s 1921 book), he starts a new paragraph with the following sentence: “Rivière described the origins of the women in Renoir’s picture, and named the male friends, but then he gave the false impression that Renoir painted an ordinary Sunday dance of Montmartre residents.” This sentence mixes together the more general kinds of information given in Rivière’s 1877 article (social origins but no names given for the women, and the impression of an ordinary Sunday dance) with the rather more detailed sort of information found in the 1921 book (where the names of both the men and women are given, along with the circumstances in which the picture was actually painted). The reader, in short, has to go back to the original sources and do a certain amount of untangling in order to determine exactly which one Herbert is using when.

Herbert’s present justification for having supposedly done what in fact he has not done—i.e., made the precise nature of his historical sources clear—is strained and circular. And in any case, since he is supposed to be writing about how pictures are seen in their social setting, he should be more specific about which social setting he is talking about, that of 1877 or that of 1921. Nor can he get off the hook by saying that Rivière in 1921 “unwittingly” undermined the myth that he and Renoir had created in 1877. In 1877 the youthful Rivière was defending and promoting the radically original painting of a rather young and unestablished artist. In 1921 he was reminiscing about his friendship with a world-famous and already dead master. The point that Herbert seems to miss is that people’s minds and motives change over time, as Rivière’s did between 1877 and 1921.

One would have thought that a writer who wanted to consider the social and historical background of such a picture would have been predisposed to compare, rather than to gloss over, the revealing differences between Rivière’s two texts. This, it seems to me, is part of what historical writing is all about.

This leads me back to why I found so much to criticize in so informative a volume by an author whose previous work I have so much admired. As I thought was clear from my review, my doubts are not about “social art history” but about the way Professor Herbert practices it in this book.