‘With a speed that still seems amazing’, wrote Clement Greenberg some sixty years ago, ‘one of the most epochal transformations in the history of art was accomplished’: the arrival of abstract art. This was, in the eyes of Greenberg and many others, not simply one new possibility added to the rest, but the one that would inevitably come to dominate: an art uniquely answerable to ‘the underlying tendencies of the age’. Today, the transformation seems less epochal. Artists, critics and theorists are more likely to point to Marcel Duchamp’s discovery of the readymade than to the advent of abstraction as the really amazing transformative leap in art in the early twentieth century. Good abstract painting and sculpture is still being produced, it might be conceded, but only in the same way that, for Greenberg, ‘good landscapes, still lifes and torsos will still be turned out’ under the reign of abstraction. Step into the contemporary wing of any museum and you will see far more representational art, in the form of photographs, videos and installations—or even more or less traditional representational painting—than you will abstraction.

Has the great adventure of abstraction faded with the century that gave birth to it? If so, then we should at least be able to get a historical grip on the phenomenon. We are still far from having such a history at our disposal, but one might hope to find a contribution to it in a book with a title like Pictures of Nothing: Abstract Art Since Pollock. The volume itself is a glossy, heavy-weight production, with copious illustrations of the many works discussed. Its text has been transcribed from the 2003 A. W. Mellon Lectures given at the National Gallery of Art in Washington by the former chief curator of the Museum of Modern Art, Kirk Varnedoe. There is thus considerable experience behind these lectures, and considerable institutional weight; the annual Mellon Lectures have produced some landmark, and often very popular, contributions to the literature of art, from Kenneth Clark’s The Nude and E. H. Gombrich’s Art and Illusion in the 1950s to Arthur C. Danto’s After the End of Art and John Golding’s Paths to the Absolute in more recent years.

Varnedoe himself was an energetic and highly influential figure in the American art world. A charismatic lecturer (and rugby player), he was born in 1946 to a wealthy family in Savannah, Georgia; his grandfather had been the city’s mayor, his father a stockbroker. Varnedoe was educated at Williams College, which has produced an unusual number of today’s prominent museum curators and directors (in his foreword, the National Gallery’s director, Earl A. Powell iii, recalls their encounters ‘in the undergraduate classrooms and on the muddy playing fields’ of the college); at Stanford, under Albert Elsen; and in Paris, studying Rodin. An early contribution was ‘The Ruins of the Tuileries, 1871–1883’, a social-historical account of Impressionism centred around the charred remains of the royal palace after the Commune. Varnedoe began curating while teaching in New York in the 1970s. His early exhibitions suggest an eye for undervalued artists and schools, such as the lesser-known Impressionist Gustave Caillebotte, or fin-de-siècle painting from Scandinavia. His collaboration with William Rubin at moma produced the contentious ‘Primitivism’ exhibition in 1984.

As Rubin’s chosen successor, Varnedoe was named chief curator of painting and sculpture at the Museum in 1988. He opened the new decade with the massive ‘High and Low: Modern Art and Popular Culture’ (organized in collaboration with Adam Gopnik, who contributes a preface to Pictures of Nothing) and continued with major retrospectives of such post-war American artists as Pollock, Johns and Twombly. In 2001, already ill with cancer, Varnedoe stepped down from moma to take up a post at Princeton, where he devoted himself primarily to preparing the Mellon Lectures. Varnedoe must from the beginning have intended these as a summation of sorts. There is for this reason a terrible poignancy to what would otherwise have been the standard disclaimer issued at the start of the last lecture in the series, when he says, ‘I am so painfully aware of how much has gone unsaid, and how much I would still like to say’; all the more so as he then goes on to quote Rutger Hauer’s words in Blade Runner—‘All those moments will be lost in time, like tears in rain. Time to die.’

For all that, there is nothing of the valedictory about Varnedoe’s account of abstraction. He urges us to forget the ‘wearisome habit’ of always seeing ourselves ‘at the end of something rather than possibly near the beginning’, the way ‘so many scholars and critics—and artists too—are accustomed to thinking that the party was over before they got there, and that everything has to be described in terms of the ruination of a former set of ideals’. In fact, Varnedoe does not much like the extremism of abstraction’s beginnings either, and seems most comfortable in media res. At the beginning of his first lecture, he explains that he will be speaking about the abstract art of ‘the last fifty years or so’ and that the payoff to this discussion would be an answer to the question, ‘Why abstract art?’ But there is something odd about the promise to offer an explanation for the existence of abstract art that has little or nothing to say about its originators—about what artists as geographically far-flung and ideologically diverse as Piet Mondrian, Kasimir Malevich and Arthur Dove might have had in mind when they began painting abstractly in the second decade of the last century. Varnedoe has little sympathy for the origins of abstraction; these amount in his view to ‘a culture of crypto-religious, timeless certainties, associated closely with the new monolithic collectivism in society’—a breathtaking simplification that allows him to dismiss both the theosophical and the revolutionary aspirations of the pre-World War ii abstractionists at once.