Delicious dips into the vortex

Sir Michael Levey has read every number of The Burlington Magazine from the past 99 years to make this provocative, sometimes malicious or perhaps just robustly mischievous selection of 44 articles. The result is a rough gallop across the wildest stylistic century of the Christian era, from a faux-naif 1903 account of Lizzie Siddal by Dante Gabriel Rossetti's brother, William Michael, to a ruthless 2002 demolition of David Hockney's the ories on optics and the Old Masters by David Bomford.

None of the essays is too long; many are brief and easily absorbed; few are predictable. Roger Fry, Herbert Read, Augustus John, E. H. Gombrich, Kenneth Clark, Rudolf Wittkower and Hugh Honour are among the contributors; this is the world of art scholarship, armed and combative, lively and disturbing, aimed at the reader who can cope with controversy.

There is Brinsley Ford's sharp-edged obituary of James Ensor, published in error when Ensor was alive and serving in his Ostend shell shop. In the same mood, Levey selects Erwin Panofsky's wonderfully convincing 1934 analysis of the Van Eyck Arnolfini double portrait, which Lorne Campbell subsequently subjected to - and I quote the impish Levey - "a detailed devastating analysis and rejection with its erudition impugned". The Burlington has been called "that stuffy magazine", but not on this showing.

Levey has reservations about Fry, a contributor on impressionism. But impressionists are old hat, outdated by Cezanne, so instead of a predictable French rapture, Levey has selected Fry's 1910 article on "Mohammedan art" proving that Islamic art was entirely parasitic on the art of the Persian Sassanid dynasty, which was established centuries before the birth of the Prophet.

Another shrewd thrust at an established reputation is Augustus John's rollicking encomium on his sister Gwendolen, illustrated by four of her paintings of anorexic, suffering young women in monochrome and one scornful self-portrait of Gwen in a scarlet dress. Her brother recalls that she painted for choice in a dark cellar among her beloved cats and once posed, scantily draped, for a colossal female statue that Rodin was carving to commemorate Whistler. When this was completed bar one arm, the commmitee had second thoughts and bought a replica of the Bourgeois de Calais instead. Gwen herself painted, "with slight variations", nuns, children, flowers and cats. John concludes solemnly: "Here was the greatest woman artist of her age, or as I think, of any other." The end came "when, feeling the need for a change of air, Gwen John took the train to Dieppe. She collapsed on arriving. She had brought no baggage whatever, but as it turned out had not forgotten to make in her will a suitable provision for the cats." With a brother like that, who needs enemies?

What is so impressive about The Burlington is that it has always been able to stand back from its contributors, let them have their persuasive say and then publish exactly the opposite views in the next issue. Successive editors must have found it painful to resist the demands of loyalties and commitments, yet The Burlington has remained pragmatic throughout a dangerous century - a broad church, but doing all the voices. Here they are, the manipulative voices of the wordsmiths, whose methods are at least as interesting as the causes they seek to project. I am a lecturer in architectural and garden history somewhat isolated in a university department of history of art and intrigued by the difference between my voice, my methodology, vocabulary and aesthetic philosophy, and the voices of my colleagues next door, who teach fine arts, painting and sculpture. Because buildings and gardens are closely related to functioning human life, the analy sis and judgement of them is reassuringly weighted to the objective. A house or a church has no need to justify its existence; it has only to create an ambience appropriate to its function. Similarly a garden has to grace or exclude a landscape. One of the articles in this selection, J. B. Bullen's "Sarah Losh: architect, romantic, mythologist", published in 2001, exemplifies perfectly this objective critical language that architecture demands. This is an example of crystal-clear analysis and wide ranging scholarship. It presents Losh, her friends, her environment, her life and her buildings without a single sentence of obscurity.

Art is different, art is subjective not objective. Therefore, art needs to be sold, but after reading these fascinatingly variable sales pitches, I begin to sympathise with the quite different voices required to project that subjectivity. Ever since photography was invented, art has been in both retreat and advance: retreat from the figurative, advance in claiming to be more real than reality. Art needed faith and disciples ready to make the leap, which a faith always requires, across mere logic - hence the seductive voices. The Burlington kept its head when all about it were losing theirs and blaming it on Fox Tal bot. It offered a prestigious, beautifully ifius trated forum, establishment in feeling but prepared to take risks with the rebels. Its contributors, the preachers of its multiple faiths, deployed two voices, one of honest scholar ship, the other of the vortex - that critical trick where words, syntax and ideas swirl impressively together to bypass clarity and hypnotise the reader.

Scholarship's voice can come in variable pitches. Svend Eriksen researches away to find exactly when Louis XV's arts minister, the Marquis de Marigny, abandoned the rococo for neo-classicism simply by checking his picture frames. Benedict Nicholson, the magazine's longtime editor, relies on bluff, bombast and name-dropping to analyse the 1952 "Caravaggio and the Netherlands" exhibition. "It would have been instructive," he growls, "to have hung Caesar van Everdin gen alongside Bor and Bronkhorst, and Jacob van Oost and the still-puzzling Finson alongside Seghers and Rombouts. Couwenbergh could perhaps have been more impressively presented." Such de haut en bas treatment gives art a bad name.

Martin Davies, in contrast, dissects Madame Moitessier's anatomy with such helpful expertise that Ingres would have been delighted to learn that he had been so subtle: 'The second curve is the short line of the waist. The third (but from left to right, by a form of contrapposto giving stability to the whole) is the upper line of the dress across the shoulders, continued behind the forehead in the glass; the table's twirl being repeated (the strata earthquake-wrenched) by the right arm." On second thoughts, this is not so much scholarship as a vortex, though a vortex-was not really needed in this case as the picture sells itself without verbal projection.

Vortices are needed only to carry mysteries, to convey significances that are beyond rea son. At the end of his brilliant, definitive analysis of Cézanne's work, Maurice Denis desperately needed such an expression of supreme admiration for the artist's rendering of fruit in his still lies.

But instead of creating a vortex he falls back on one word "classic" - which is quite inadequate to convey what he believes Cézanne has achieved: an object in its own right, the platonic essence of apple or orange, not a mere reflection of a reality.

Denis was writing in 1909. David Sylvester, writing on "The evolution of Henry Moore's sculpture" had, in 1948, a harder challenge, but better tactics. He began by laying down Moore's four distinct periods, then offered a lifebelt of interpretation that anyone could grasp: convex forms are passive and restful, concaves are active and dynamic, a simple satisfying paradox, though possibly untrue. All he needed then were a few vortices to carry a reader's reason to an O Altitudo , as in the elm-wood Reclining Figure of 1939. This was, he wrote: " A rhythmic expression [note the impressive italics] of female sensuality (that would wake most readers up)... cathedral and cave-like in its mysterious depth and yet again more than this, for we need impose no arbitrary associational limit on the life embraced by its mercurial sensual rhythms, its stress of mass against mass, its grandeur and weight, its play of light that now illumines, now gives way to shadow, now reappears, as the holes in the flanks, and the dense masses between them serve as win dows and wall."

If T. S. Eliot could make the Four Quartets sound gnomic by antithesis, why not Sylvester? In that 1910 cubist phase, which he wisely abandoned a year later, Picasso was more of a challenge. But with Portrait of Daniel-Henry Kahnweiler to project, Roland Penrose first softened readers up by reporting that a four-year-old child, presented with the equally impenetrable cubist Portrait of Ambroise Vollard, had instinctively cried out: "That's Monsieur Vollard." Shamed by such infant perception, we read on. When the rhetoric of his vortex becomes too remote from reality, Penrose slips in an easy concrete reference, the Alhambra ceiling, for reassurance before plunging on again: "The body from the shoulders to the hands, is again composed of planes which can be taken to protrude or retreat without finally destroying the total mass and allowing us the delight of letting the eye wander in and out of its successive crystalline passages. The organisation of the surfaces forms a pattern that recalls the intricate honeycomb design of the Moorish vaulting in the Alhambra. Here, although it floats in the vertical plane of the picture, it is like ripples on the surface of water. The imagination is stimulated to find in it scenes which though they are ambiguous appear undoubtedly to exist. It is excited by the rhythmic life of this new reality and devises with delight its own interpretation." Though Penrose's vortex is a frank admission of subjectivity, I still find myself both seduced and enriched. Difficult art demands intoxicating writing. Ruskin showed us the way, and what is life without a broadening sensibility, or appreciation without the occa sional leap in the dark?

This book supplies many such leaps, then adds refreshing malice. In a rude editorial gesture across the Atlantic, "Upon a peak in Brentwood" is an appraisal of the Paul Getty Center outside Los Angeles. It is in the wrong place, it is a dull building by the wrong archi tect - Richard Meier instead of Frank Gehry - but it is otherwise perfect. A sublime evening view of the loutish complex, a geometric scrabble against the serene Pacific, clinches the point visually. Art criticism is alive and well and still sharpening its teeth in London.

Timothy Mowl is reader in architectural and garden history, University of Bristol.