Men of different generations and presumably social worlds, Anthony Powell and Craig Raine aren't much alike as writers. But the novelist's Miscellaneous Verdicts and the poet's Haydn and the Valve Trumpet are both very good, solid selections of occasional writing. The five hundred pages to which they both run are mainly literary journalism, with some illuminating essays on the social-historical from Powell, and vivid side-glances at painters and painting from Raine. With all their differences, the two writers have one thing in common. Both dislike most kinds of academic literary criticism. And this antipathy can't be disentangled from the effective virtues of their work.

It seems safe to assume that academics have as much right to discourse on books as have poets and novelists to write them. Nor do minds as able as Powell's and Raine's need telling that in modern society the arts depend on a current of ideas which it is the universities' task—at least in theory—to provide and protect. The trouble comes with the theory.

Nobody could pretend that universities are at present, or were ever, especially alive with applied intelligence. In addition, we are in a difficult phase of academic literary criticism, which has the air of getting cleverer and cleverer while simultaneously moving close to pointlessness. The new quasi-theoretical modes as often as not find a use for Shakespeare or Jane Austen or T. S. Eliot by exposing them, morally or politically or otherwise, as no good. This is annoying for writers and farcical for readers.

Though sometimes plainly motivated, this effect is basically incidental. Academic life is now governed by the thesis; and a thesis is required to show an authoritative mastery of its literary subject easily converting to a stance of superiority on the part of the researcher. Moreover, such research techniques have managed almost universally to demand of all literary criticism that it have what is referred to as 'system'. This seems reasonable. Unfortunately, what passes for system academically is often no more than mechanism, producing results painfully shallow in comparison with the real systems of high-powered human intelligence.

Defending literature now can place the liberal academic in positions which it's not altogether ludicrous to relate to that of the trapped liberal of the Thirties, confronted by competing totalitarianisms. And those positions are inherited by university-trained writers like Powell and Raine. Both enunciate principles as congenial as they now sound dated: Powell's civilised 'plea for mutual tolerance among authors writing on the same subject', Raine's brave 'nothing is more difficult than being open-minded.' Any liberal reader reads and admires and sympathises with their impassioned defence of the writer as against the academic. And of course poems and novels are better and more vital than critical essays. But at the moments when both try to find a way of saying why this is so, they get curiously trapped between the philistine and the Romantic-aesthetic. Thinking about the arts is at once more important than they sometimes make it sound, and harder….

Powell and Raine come together in a peculiar and very interesting Britishness, their sense of art an English one. This is what makes both, for all their late-Romantic aestheticism, simultaneously risk the philistine in leaning backwards towards an Augustan lack of cant. (Johnson's often misunderstood attack on the corruptive power of arrogant pretension, 'No man but a blockhead ever wrote but for money,' is the great text for journalism. And both Powell and Raine can take on a special sharp-edged no-nonsense realism which shows them as Sons of Sam.)…

Miscellaneous Verdicts is itself 'all right' in any number of different ways: its greatest distinction may be the dance which Powell made his reviewing perform for years around the idea and the fact of a writer's rectitude. Craig Raine's Haydn and the Valve Trumpet is as gifted and richly entertaining as Powell's volume. But it offers one contrast in style so marked as to be almost ideal. Raine isn't, in a sense, concerned with being 'right'. In fact, he reveals that if a critic is good enough, he can afford to be wrong.

Being wrong is his general theme. A fine brisk essay on Joyce ('New Secondhand Clothes') surveys an earlier stage of the current battle over Joyce texts by stating the principle that misprints occur and don't matter. This is true up to the point that meaning is more important than text. But Raine gives his theory more space in his opening essay. He mentions a critic who recently made the mistake of arguing that Haydn was influenced by a form of the trumpet which proved not to have been invented until after the composer's death. Raine's point is that, in this case as everywhere else, it's the music that counts, not the nonsense we talk about it. Hence Raine's choice of a title for his book which works by a kind of triumphant wonkiness. Interestingly different from the amused offhand anonymity of Miscellaneous Verdicts, the attractively nubbly Haydn and the Valve Trumpet is a short Raine poem in itself.

Raine's tough commonsensicality, his respect for real life and for the serious 'game' which he takes poetry to be, and most of all the intelligence of his good-natured gusto and rage, all work-together to give integrity to his arguments. Yet in the simplest possible way he can get things wrong in a manner that may dent his case just a little. The dust-jacket of Raine's book sets his title in a box against a fine etching by Rembrandt one perhaps even too fine to have been used to sell a book ('there are perhaps worse places to read about Melancholy than a publisher's office'). The inside of the jacket identifies this beautiful image as The Young Christ Disputing with the Doctors. This is a naming which Raine defends in one of his later essays, 'At a Slight Angle to the Universe', which rebuts sentimental linking of the artist with the child. The artist may be childlike, he suggests, only in his or her capacity to correct stale and sedate quasi-philosophies, like the young Christ in his dispute with the Doctors of the Temple; and Raine turns to the etching by Rembrandt usually known as Joseph Telling his Dreams, claiming that its true subject is that of Christ with the Doctors. If Raine had been right, there might have been even less to be said for using it as a dust-jacket: identifying critics with Jesus just must be a mistake. But luckily Raine is wrong. The subject is what it has always been taken to be, 'Joseph Telling his Dreams'.

Rembrandt left behind at least two real treatments of the topos of Jesus with the Doctors, in each case leaving the subject iconographically unmistakable: brief but definite indications of monumental masonry show that the location is the Temple. The print on Raine's cover is no vast stone edifice filled with Scribes and Pharisees: it is an intimate domestic interior. The old lady behind the boy is in a bed, perhaps a day-bed—you can see bed-curtains, not to mention a night-cap; she is conceivably Joseph's mother, Rachel, who bore him very late in life (though she was actually dead by this stage of Joseph's existence). The figure surely can't be the young Mary, mother of Jesus, and she wouldn't lie around in a Jewish temple anyway. The loving old man on the left isn't a Pharisee but Joseph's adoring elderly father, Jacob; the sullen averted faces to the right aren't intellectuals but his embittered older brothers, soon to attempt his murder in jealous rage. And the wonderfully intent boy at the centre has the face of a poet, not of God; he is dressed not in sanctity but in the very best and most expensive possible 17th-century boy's topcoat—the many-coloured dreamcoat, in short.

The mixture of great virtues and great mistakes is an essential part of what Raine is doing and saying. At one point, he lays down the sturdy affirmation, 'Eliot is a poet by whom critics are judged'—and this is certainly true. But Eliot was a critic himself, as many poets are. He goes on to argue that the general theme of Eliot's verse 'from first to last' is 'Live all you can. It's a mistake not to.' Even Henry James, whose novel The Ambassadors is the source of this phrase, wouldn't have said it in propria persona: it's odder still from Eliot.

Raine is a splendid critic of the textures of language, the 'pidgin' or 'Babylonish dialect' that each artist makes his own. On Dickens, on Joyce, on Elizabeth Bishop and John Betjeman—perhaps the best essays—he has things to say both brilliant and new. But he wouldn't have said them, paradoxically, had he not been a critic capable of mistakes. In all his essays he brings virtues easy to class as 'journalistic' up to the level of the genuinely literary. He does so from a strong refusal to cut the arts out of life. 'Poets hate the sanitised, sentimental, overly spiritual version of what they do. They always want the unpoetical.' And: 'If there isn't the sustained effort to accommodate the unpoetical, poetry is likely to revert to the poetical.' This is a poet speaking, a voice too individual to be mistakable for Jesus. But Joseph is quite good enough.