‘What is a ghost?’ Stephen Dedalus asks in Ulysses, and promptly answers his own question. ‘One who has faded into impalpability through death, through absence, through change of manners.’ Not a figure who is entirely unreal, just one who has become a little faint, lacking in physical immediacy. Perhaps someone who lives in the memory only, not an inconsiderable form of life after all. Or in possibility, a spirit from the future.
When they are not insisting on the absolute non-existence of ghosts, reasonable people are apt to make rather nervous jokes about them. Asked if he believed in ‘ghosts and apparitions’, Coleridge said no. But that wasn’t all he said. The complete sentence was: ‘No, madam! I have seen far too many myself.’ Quoting this passage, Susan Wolfson goes on to remind us that Freud comments on a neat variant of the same gag: ‘Not only did he disbelieve in ghosts; he was not even frightened of them.’ As Dr Johnson said on the subject of the possibility of a post-mortem appearance among the living (he too is quoted in this remarkable book), ‘All argument is against it, but all belief is for it.’ And not just belief. History has a role as well. The superstition, if it is one, had been around for five thousand years, Johnson said; when Byron quoted Johnson in Don Juan, he upped this to six thousand, concluding that ‘Whatever bar the reason rears/’Gainst such belief, there’s something stronger still/In its behalf, let those deny who will.’ The point perhaps concerns not so much the questioned reality of ghosts, or their undoubted persistence in the imagination, as the trouble they cause for reason. It is because they don’t exist in several important senses that they do exist in others. The distinction between the living and the dead matters; we can’t do without it. And yet there are so many ways of crossing the gap that the mythological migrations, the metaphors that are more than metaphors, are not likely to go away.
Literature is full of examples, which may be especially instructive when they don’t concern literal ghosts. ‘It is impossible to lay the ghost of a fact,’ Marlow says in Lord Jim, a little before Stephen Dedalus starts his discussion in a Dublin library – only four years before, if we think of the 1904 setting of Ulysses rather than the time of the writing of the remark. I’m pretty sure Marlow is right about facts, but do they produce a particular form of ghost? And what do we think of all those unfacts that are equally hard to put to rest? They come in many varieties, sinister and comic, and distinguished modern poets especially have been very keen on remembering what they haven’t done. I’m thinking of Robert Frost’s ‘The Road Not Taken’ and the unrealised walk in the first part of T.S. Eliot’s first Quartet.
Frost says ‘Two roads diverged in a yellow wood’, and recounts his regret that he ‘could not travel both/And be one traveller’. He looked at one road, then took ‘the other, as just as fair/And having perhaps the better claim,/Because it was grassy and wanted wear’. The better claim rests on apparent neglect, and the allure for the traveller of belonging to the few rather than the many, but Frost immediately cancels any empirical basis for the distinction:
Though as for that the passing there
Had worn them really about the same,
And both that morning equally lay
In leaves no steps had trodden black.
The poem is not really about the road the poet took, or even about the titular road he didn’t take: it’s about the story he will tell and the difference the non-difference will make. Still, both roads are there, and both are needed:
I shall be telling this with a sigh
Somewhere ages and ages hence:
Two roads diverged in a wood, and I –
I took the one less travelled by,
And that has made all the difference.
As Wolfson says, the road not taken ‘may be counterfactual, but it is not counter-actual’.
The Eliot instance is even more baffling.
Footfalls echo in the memory
Down the passage which we did not take
Towards the door we never opened
Into the rose-garden.
Here there is only one road but the not taking it has become part of a perception rather than a story. There is an aural memory of a non-event, of an actual failure to take the real passage to the real door into the real rose garden. The recall seems curiously authentic – how could we not hear what we keep hearing? – and blatantly faked at the same time: if he didn’t go, he didn’t go. But then Eliot adds: ‘My words echo/Thus, in your mind.’ The analogy, almost an allegory, is clear enough in one sense. The poem talks to us as if it were made of steps we did not take, and indeed it is. Every poem is some sort of fiction for the reader. The poet took the steps, whatever they were, or pretended to. But Eliot’s example doesn’t just say that the words of the poem are fictional or personal, it says they are an echo of a non-echo, the voice of a ghostly second self, someone who keeps not taking the road not taken. Is this what a poem is? What would it mean to think so?
Romantic Shades and Shadows situates itself very clearly in the current academic landscape of literary studies. The long reign of the so-called close reading associated with the New Criticism founded in the 1940s – the contemporary alternative to it being not another way of reading but an interest in society and history rather than ambiguity and irony – was severely challenged in the 1970s on grounds of its exclusive attention to ‘the words on the page’, and the static, gentlemanly nature of its canon. Wolfgang Iser’s The Implied Reader (1974) and Stanley Fish’s Is There a Text in This Class? (1980) took reading in different directions, and the New Historicism was soon in full swing.
Some of us never gave up what we thought was close reading, but it is only quite recently that the label itself has returned to use through quasi-musical variants on its wording. ‘Surface Reading’, an article by Stephen Best and Sharon Marcus, appeared in the journal Representations in 2009, and argued that much close reading was so eager to dive into linguistic implications and secrets that it couldn’t see what wasn’t hiding in plain sight. Franco Moretti’s Distant Reading (2005) recommended an alertness to data and difference, and a certain aloofness from stylistic analysis. There is also the method of ‘too close reading’ wittily practised by David Miller, especially in his book Hidden Hitchcock (2016).
Wolfson’s work is not unrelated to this last example – both writers are concerned with seeing apparitions and being glad to see them – but she tends to think that patient, uncensored curiosity will do the job that Miller entrusts to critical obsession. Returning to the old mode of close reading, Wolfson declares a direct allegiance to Cleanth Brooks’s idea of ‘the pressure of context’, where the words of a poem represent possibilities of meaning rather than the straight delivery of a preconceived thought. Even so, her writing, attentive to every twist and turn of the words on the page, has a riskiness about it that we don’t associate with the New Criticism. Unless of course we count William Empson as belonging to the school despite himself (he hated its axiomatic claim that ‘the design or intention of the author is neither available nor desirable as a standard for judging the success of a work of literary art’) since his early criticism is indistinguishable from the idea of the long shot in the world of meaning, the semantic gamble no one else would think of making.
The desire to ‘speak with the dead’ became a mantra for New Historicism as a critical method. Sometimes this meant speaking for them, or getting them to confess, and Wolfson evokes an interesting ambiguity in a remark by Paul de Man and its citation by Marjorie Garber. De Man writes of ‘making the death speak’, which in Garber’s quotation becomes ‘making the dead speak’. As Wolfson notes, the quotation has its logic and perhaps the original was a misprint. But then de Man kept his phrasing in a reprint; and Garber kept hers too. Who was speaking?
This is not quite Wolfson’s question, which is all about hearing. She wants to listen to the dead, and for the dead, to catch their murmurings even or especially when they look like accidents of language. ‘This is a book about spectral linguistic agencies,’ she says; about ‘chance associations, ruptures of logic, figural recurrences, and overproductions’. The author doesn’t have to intend these ghostly effects, and certainly not the way the reader finds or interprets them. In this perspective the author is not dead as Roland Barthes thought he or she had to be if the reader was to be liberated (‘The birth of the reader must be at the cost of the death of the Author’), but dismissed, let’s say, sent off for a walk while the unimpeded reader looks around among the scatter of meanings, plausible and implausible. Decisions about what to do with the multiplying ghosts, and which ones to believe in, and how, can come later. However, Wolfson does quite often discreetly call back the author, suggesting that he or she may know quite a bit more about apparently accidental hauntings than we suspect.
Romantic Shades and Shadows, after an introduction ‘setting the stage’, continues with a chapter on the loading of ostensibly unloaded words in Wordsworth (the echo of word in Wordsworth gives part of the game away). Next comes a chapter on Hazlitt’s ghostly re-creation of the figure he once knew as Coleridge (the man is still alive but the former radical figure isn’t). A further chapter considers Shelley’s ‘political poetry of 1819’, especially as it relates to the Peterloo Massacre. The massacre took place at St Peter’s Fields, and the Manchester Observer coined the name, marking, as Wolfson says, ‘the alliance of domestic brutality with the cant of Waterloo triumphalism’. When Shelley writes of ‘heroes of unwritten story’ he knows the story – of the success of protest in England – may never be written, but knows too that it takes a single letter to turn words into swords. The critical point is that such verbal overlappings may be interesting (or not), but do not prescribe or constrict interpretation. We could see the small spelling step between ‘word’ and ‘sword’ as an ironic marker of the large distance between the two, as Auden does with the words ‘reader’ and ‘rider’ (in the poem ‘O Where Are You Going’), just a vowel sound apart in language, worlds away from each other in Auden’s view of things. With Shelley, an opposite move would be better: the verbal closeness could be a promise or a hope, but one more at home in the realm of incantation than that of political practice.
Wolfson’s chapter on Byron shows how brilliantly a sceptic can put fantasies to work, and suggests that Byron was half-parodying himself even before he had constructed an image to parody. It’s worth quoting, as she does, a passage from The Siege of Corinth (1816) that makes memorable ghosts without any immediate recourse to the supernatural. The prompt is a dream-vision of a woman the dreamer doesn’t know is dead. Her eyes are still, ‘as the eyes may seem/Of the restless who walk in a troubled dream’. But the ghosts Byron describes are an impeccably natural phenomenon:
Like the figures on arras, that gloomily glare
Stirred by the breath of the wintry air,
So seen by the dying lamp’s fitful light,
Lifeless, but life-like, and awful to sight;
As they seem, through the dimness, about to come down
From the shadowy wall where their images frown;
Fearfully flitting to and fro,
As the gusts on the tapestry come and go.
There is something like the reverse of Eliot’s footfalls here: real woven figures moving in the wind, a wall full of ghosts and no actual ghost we know about yet. Wolfson writes of ‘an awful sheen of presence in the material world’.
The book closes with what Wolfson calls ‘a short story about my own haunts of reading’, but before that there is an extraordinary chapter about Yeats’s failed scheme of denial of Keats. ‘I cannot write on Keats,’ he said when asked to contribute to a memorial volume to be published in 1921. ‘I have not read Keats during the past five years, and I should have to fill my mind with him.’ Was he influenced by the ‘unlucky’ visual rhyme in their surnames? Probably not, but it can’t have helped. Wolfson shows persuasively what the problem was. Read properly, Keats was too modern and therefore an intrusion into Yeats’s project – or would have been an intrusion had Yeats allowed it. Yeats could hardly be one of the last romantics, as he claimed to be – that is, one of the genuine, conflicted modernists – if a first romantic was already as belated as he was.
The condescension towards Keats of one of Yeats’s characters is well known. This is the speaker Ille in the poem ‘Ego Dominus Tuus’:
I see a schoolboy, when I think of him,
With face and nose pressed to a sweetshop window,
For certainly he sank into his grave,
His senses and his heart unsatisfied;
And made – being poor, ailing and ignorant,
Shut out from all the luxury of the world,
The ill-bred son of a livery-stable keeper –
I don’t think anyone ever considered Keats ‘ill-bred’, since this is a matter of behaviour, and Yeats elsewhere (‘not perfectly well-bred’) corrects this error, but only at the cost of replacing injustice with snobbery. Fortunately Yeats knew in the end how to let himself be haunted, and the two poets met in their complicated love and suspicion of artifice, delighting, for example, in a Grecian urn and a golden bird. There is after all ‘a latent Yeats in Keats and a latent Keats in Yeats’. All writers invent their precursors, as Borges said, but the precursors too can be part of the invention.
There are treasures in all of these chapters, but the tour de force, the greatest adventure, to use Wolfson’s term, and the set of interpretations that will drive some readers crazy, lie in the essay on Wordsworth’s wordplay. Some of us have been set up for this game by Nabokov’s Pale Fire, where Wordsmith is a college resembling Cornell, and Goldsworth a judge whose house our unreliable hero is renting. We see how pieces of writers’ names can be switched, creating a kind of literary quilt, and also that names, if they are made up of ordinary terms (word, smith, gold, worth), may revert without warning to their ordinary life.
Wordsworth doesn’t juggle with his name as other writers do: Shakespeare, for example, has plenty of fun with will, and Donne, speaking of the forgiveness of sins, writes: ‘When thou hast done, thou hast not done,/For I have more.’ But the fact that Wordsworth doesn’t play games of this kind doesn’t mean he doesn’t play games, and in Wordsworth’s Fun, a book recently reviewed in the LRB (4 July, 2019), Matthew Bevis persuasively suggests that Wordsworth had a sense of humour that has escaped almost all of his readers. He can’t not have noticed how easily his name falls apart into separate words or phrases: will, I am, words, worth. And even if he hadn’t noticed, others were prepared to make jokes using his name. Coleridge suggested, when sending a copy of his poem ‘The Nightingale’, that ‘you’ll tell me what you think my Bird’s worth.’ And Byron, less politely, liked to refer to the poet as ‘Turdsworth’.
Wordsworth keeps ‘his proper name off the main turf of The Prelude’, Wolfson says, ‘sounding it only in syllables tuned to other missions’. But then those other missions have interesting turfs of their own. What happens when we read ‘The calm existence that is mine when I/Am worthy of myself?’ Or the passage that started Wolfson off on her reading of ‘name-haunts’. In a dream an Arab shows Wordsworth a stone and a shell:
‘This’, said he,
‘Is something of more worth’; and at the word
Stretched forth the shell.
Word and worth in quick, if reversed sequence. There are many more examples, as perhaps there must be, given the commonness of the words we find in Wordsworth’s name. Could he really abjure them? He was the poet who called his mind a ‘strange rendezvous’. Certainly, if we wish, we can dismiss these apparitions as accidents of meaning, simple consequences of the fact that, like Eliot’s Sweeney, we gotta use words when we talk to people. Or we can wonder where we are, as J.H. Prynne, among many others, did when faced with the case of Ferdinand de Saussure, who busied himself finding unsuspected anagrams in Latin and other poetry. (Lévi-Strauss called Saussure’s readings ‘tinkering’, while Jean-Jacques Lecercle said they were ‘demented’.) Do these kinds of discovery prove the existence of, to borrow Wolfson’s wording, ‘an apparently motivated system’ or a ‘too apparently motivated reader’? Prynne’s question is ‘would the effect merely imitate the appearance of significance, or might the work of such selection in fact have installed significance?’ We have arrived at the grand old question of over-reading, or reading meanings into texts. Meanings that aren’t there, that is. But where is there, and who is to tell us whether we found our ghosts or invented them? There are conventions in such matters, but conventions can be weary and obtuse as well as helpful.
An anecdote and a return to Nabokov, always a friend when linguistic games are an issue, may help us to focus a little. He wrote a short story called ‘The Vane Sisters’, which ends in an acrostic. If we read the first letters of the words in the work’s last sentences, we get a message from two dead women in the narrative, a ghostly communication that the author of these sentences can’t read: ‘Icicles by Cynthia meter from me Sybil’. The New Yorker turned the story down because it didn’t go in for such fancy stuff. Long ago, trying to write about this story, I thought of Saussure’s anagrams, and the interest of Nabokov’s narrator in acrostics hidden in Shakespeare’s sonnets. He finds the name of President Taft – twice. Rather idly checking these discoveries, I was pulled up short by Sonnet 14, which actually, by this method, begins to name Nabokov:
Not from the stars do I my judgment pluck
And yet methinks I have astronomy;
But not to tell of good or evil luck,
Of plagues, of death, or seasons’ quality.
Fortunately the rest of the sonnet abandons the game. If it had continued, though, once I had got past the idea that I was insane, I wouldn’t have wondered how Shakespeare managed to foresee the coming of Nabokov, I would have registered an eerie reminder that the time of reading is not linear. And I would have been glad of the confirmation that accidents of language talk all the time. The question is whether we can make anything of what they say.