What Henry Knew
If you tasted it, it would first taste bitter,
then briny, then surely burn your tongue.
It is like what we imagine knowledge to be:
dark, salt, clear, moving, utterly free,
drawn from the cold hard mouth
of the world, derived from the rocky breasts
forever, flowing and drawn, and since
our knowledge is historical, flowing, and flown.
Elizabeth Bishop, ‘At the Fishhouses’
‘Like what we imagine knowledge to be’. There are many ways of imagining knowledge – this is a proposition not likely to provoke much dispute. But how about this one? There are many ways in which the imagination creates and imparts knowledge. Is this claim self-evident, or just untrue – untrue because the imagination, whatever its wonders and virtues, doesn’t deal in knowledge in any but the weakest, least demanding sense? There are all kinds of good things which are not knowledge, and we should not betray them by giving them the wrong name.
The worry about the relation between literature and knowledge is a very old one, and it’s not getting any younger. When Dorothy Walsh, in an elegant book called Literature and Knowledge, published in 1969, said the worry was old, she meant it went back at least to Plato. When Stathis Gourgouris says it is old, in a book called Does Literature Think? published earlier this year, he means the same thing:
The idea that literature might harbour its own mode of knowledge is ancient, at least as old as the so-called quarrel between poetry and philosophy and Plato’s notorious expulsion of the poets from the city in the Republic. It is fair to say that since Plato’s famous decision there has been an implicit but consistent association of the poetic act with a peculiar, mysterious, and even dangerous sort of knowledge.
Actually, even Socrates thought the worry was old, and apologised for his dismissal of poetry by saying: ‘But in case we are charged with a certain harshness and lack of sophistication, let’s also tell poetry that there is an ancient quarrel between it and philosophy.’
But Plato’s worry is not ours, and indeed our worry, in 2003, is perhaps not quite the worry we might have had, did have, in 1969. Or if the question we are asking is the same – to quote Dorothy Walsh, ‘What kind of knowledge, if any, does literary art afford?’ or, more delicately, ‘Do works of literary art, when functioning successfully as such, have any intimate engagement with what may be called knowledge?’ – our reasons for asking it are different, and so is our idea of what might constitute an interesting answer. Walsh thought that the disengagement of literature from direct knowledge claims might ‘be seen as the liberation of literature from the alien and extraneous burden of cognitive concern. So liberated, literature is free to develop its potentialities strictly as art.’ The opposite view, she suggested, was not engagement with direct knowledge but a different sense of disengagement, the view that ‘the disengagement provides the opportunity for the recognition of the distinctive kind of cognitive significance literary art can have.’ ‘Shall we see the disengagement as the liberation of Ariel?’ Walsh asked. ‘Or . . . shall we say that the magic island . . . cannot be abandoned and that the control of Prospero over both Ariel and Caliban must be sustained?’ I don’t think many people are recommending the liberation of Ariel these days, or a picture of literature ‘strictly as art’, and I don’t wish to recommend them myself. But I do want to wonder, as Walsh did, whether the only alternative is total submission to Prospero.
There is an excellent focus for the old and new question, a brilliant brief statement of its current force, and a way of holding the whole issue before our minds, in Peter de Bolla’s book Art Matters. De Bolla is looking at a Barnett Newman painting (Vir Heroicus Sublimis) in the Museum of Modern Art in New York. He has decided that the usual critical questions – what does this painting mean, what is it trying to say – are the wrong ones. He offers one or two not all that appealing alternatives (‘how does this painting determine my address to it, how does it make me feel, what does it make me feel’) and says that ‘beyond these questions lies the insistent murmur of all great art, the nagging thought that the work holds something to itself, contains something that in the final analysis remains untouchable, unknowable.’ Then De Bolla arrives at what I find the truly haunting question: ‘What does this painting know?’
The question has two immediate and very interesting implications. First, that a painting might know something that the painter didn’t. And, second, that the painting probably knows a lot that it is not going to tell us. I’m interested in the murmur of small art as well as great art – I think small art knows things, too – and I want to put the question to literature rather than to painting, but the question is the same. To frame it rather schematically, thinking of Proust and asthma, say, we could ask, not what Proust knew about asthma or what we know about asthma or what doctors knew or know about asthma in Proust’s time and in ours, but what A la recherche du temps perdu knows about asthma: what it knows and perhaps will not directly tell us, or what it knows that only novels know, or what it knows that only this novel knows. Many see dangers in such personification – the novel is not a person and can’t know anything, only novelists and readers can – but for the moment I should like the question, and the figure of speech, just to hang in the memory, like a motto, or an old tune.
My slightly frivolous title, ‘What Henry Knew’, takes us straight to Henry James, of course, and the (feeble) joke is meant, among other things, to indicate that I recognise how obvious a move this is, once we have started on the question of literature and knowledge. It was James who wrote so eloquently, in relation to the publication of Flaubert’s letters, of ‘the insurmountable desire to know’, and who thought, in that context, that ‘some day or other we shall surely agree that . . . we pay more for some kinds of knowledge than those particular kinds are worth.’ But to read these words, and to think of the knowledge at issue – the fact, as James says, ‘that the author of calm, firm masterpieces . . . was narrow and noisy’ – is to remember how many kinds of knowledge there are, how much work the words know and knowledge are asked to do, and how varied that work is.
If we read the actual sentence I have massacred for my title, for instance, at the point in the novel from which James takes his phrase, we come upon another kind of knowledge entirely: not the goal of curiosity but the fruit of experience. In his preface to What Maisie Knew, James writes of the appeal for the novelist of a child’s ‘confused and obscure notation’ of a tangle of adult relations – namely, the goings on of her divorced parents and their changing companions – and then says that it was important for him that Maisie should see more than she understood. ‘Small children have many more perceptions than they have terms to translate them.’ (I shall return to this phrase.) ‘She has,’ James says of Maisie, ‘the wonderful importance of shedding a light far beyond any reach of her comprehension.’ This is clear enough, and James’s remarks are of enormous technical interest. But the text of the novel itself is more ambiguous, and is ambiguous precisely because of the strong and multiple valences of the word know. Do Maisie’s perceptions shed ‘a light far beyond any reach of her comprehension’? Well, perhaps not beyond any reach. Just beyond the reach we are likely to assign to it. But is that the right reach? The novel is not called ‘What Maisie Saw’, or ‘What Maisie Failed to Understand’. The whole beauty of the thing is that we can’t know for sure what Maisie knows. Tony Tanner has said that ‘in a sense the book hinges on what Maisie does not know’ – especially what she doesn’t know about sex. But by the same token she can, as Tanner also suggests, deeply know all kinds of things without knowing the basic facts. Right at the end of the book the governess Mrs Wix gives Maisie a piece of information about the whereabouts of two of the errant adults. Maisie says: ‘Oh I know.’ James then writes: ‘Mrs Wix gave a sidelong look. She still had room for wonder at what Maisie knew.’
We still have room for wonder at what James knew – and in the light of my question about literature and knowledge, at what James’s novels knew. I hope my title, if viewed suspiciously enough, may help to suggest that to ask what Henry James knew is also to ask if his novels know the same things. I want to use a great novel by James to explore some of the working meanings of the notion of knowledge; and I want, at the same time, to use the notion of knowledge to open up a great novel in what I hope is a fresh way. The novel is The Wings of the Dove, first published in 1902. This novel seems to know a lot about knowledge, and especially about knowledge and words, and about knowledge and novels.
It will help clear the ground a little if we look briefly at some of the ostensive definitions of knowledge the novel handily offers us. They correspond to, but also complicate, the OED’s central definitions, especially the ones that pertain to knowledge by experience or personal acquaintance, knowledge as awareness of the facts, and knowledge as an understanding of patterns of relations or connection. For the sake of convenience, and following the tutelary spirit of William Empson, I’ve grouped the meanings of knowledge in The Wings of the Dove in seven sets.
One. In this novel, you can know that something is the case. This is the most frequent meaning of the words know and knowledge. You could know, for instance, although you don’t, what terrible disgraceful thing your father has done. In the opening pages Kate Croy is waiting for her elegant but seedy dad, and James’s prose is at its insidious best. Lionel Croy is, we learn, ‘a terrible husband not to live with’. ‘Those who knew him a little said “How he does dress!” – those who knew him better said “How does he?”’ ‘Nothing . . . was more wonderful than what he sometimes would take for offence, unless it might be what he sometimes wouldn’t.’ What’s important here is that the possibility of simple if disagreeable knowledge is mentioned only because it is so resolutely refused – a pattern which comes up again and again in the novel. ‘Whatever it was that was horrid . . . that he had done,’ the wife and daughters think, ‘thank God they didn’t really know.’ A little later Merton Densher, the man Kate is in love with, asks her just what it is her father has been up to.
‘What has he done, if no one can name it?’
‘He has done everything.’
‘Oh – everything! Everything’s nothing.’
‘Well then,’ said Kate, ‘he has done some particular thing. It’s known – only, thank God, not to us.’
The full text of this essay is only available to subscribers of the London Review of Books.
You are not logged in
 Stanford, 424 pp., £42.50, June, 0 8047 3213 2.
 Harvard, 157 pp., £10.95, March, 0 674 01110 4.
 Cambridge, 272 pp., £40, March 2000, 0 521 65230 8.