A while ago I blogged about what's going on, intuitively, when you're reading a really good book, using Hilary Mantel's Wolf Hall as an example. But, of course, many of us do read a really good book for a conscious, specific purpose. And if you have to write at length about it then you have to read even more clear-headedly. The first time I did that was for my MPhil dissertation, and the book was A S Byatt's Possession.
I was writing a novel which wasn't, then, called The Mathematics of Love, and there were things I wanted to say about my first, 1819, story that couldn't be said until a 1976 world. I balked at the all too well-used letters-in-the-attic scenario, but I invented my planning grid, and used it to track and connect themes, images, ideas and places, and one mysterious child across two wholly separate plots. Then my squirming, half-formed novel was rejected by an agent on the grounds that parallel narratives don't work. Aha! What I was trying to do had a name, had it?
With not a little sense of thumbing my nose at her I decided to write the critical paper for my Masters about a parallel narrative novel which does work. I needed novel with two stories with wholly different casts, set in wholly different eras and well-enough written to stand up to critical scrutiny: A. S. Byatt's Possession was the obvious candidate. Like many I skipped the poetry at first, while admiring the virtuosity of the ventriloquism. But all the structural things that I'd struggled with – what goes next to what, what's revealed when, how do you make it seem uncontrived? – Byatt does with such virtuosity and brio that my first reaction was, Thank God I didn't read this before I'd finished my own. And then I went back and took the book apart to write my paper.
So what does Byatt do? Well, she certainly illustrates the risk, built into parallel narratives, that most readers will enjoy one strand much more than the other. I have one friend who prefers the modern literary-detective-campus-satire strand, but most of us fall for the wonderful Victorian love "story" that's never narrated, but only put together un-chronologically, seemingly naturally, very carefully, from letters and diaries and slippery inferences in Ash and Christabel's published work. Byatt's even said herself that her heart used to sink when she realised another chunk of modern-day literary detective story was due.
And yet, like Conan Doyle's Watson, modern Maud and Roland are our representatives in the novel: Byatt's so determined to assemble the Victorian story only from its remains that we need their help to put it together. In these narrated sections the writing is authoritative, densely-textured, magnificently precise in its diction and wide-ranging in its references. You could spend a scholarly paper tracing Byatt's use of the imagery of insect life alone, or her exploration of the possessive possessedness of the biographer, or her structural use of the Melusine myth. And yet the narrative voice is detached: coolly narratorial. Only in what I came to call the documents – the letters, diaries, poems, literary criticism and stories – does the fierce subjectivity of each character-writer seem to break out and away from the narrator's own controlling intelligence and shrewd judgement.
The most interesting thing for me as a writer is just how Byatt plays those documents within that narrative. My novel had letters in it because I was fascinated by the technical challenge of expressing for the reader what a letter-writer wouldn't say. But here were documents used as clues for readers and for characters: as structure, as image, as plot and personality. Here were letters not sent or not delivered, diaries that didn't say things, private meanings in published works, "secondary sources" used as ammunition. Here was a long statement – implicit and explicit – about why people write. And here were three moments of a great love story, arguably the most important moments of all, that couldn't be put together from characters' writings, because they were beyond the characters' own words.
So what did Possession do for my own writing? Well, when people started saying that they loved the two strands of The Mathematics of Love but spent the whole novel wondering when they were going to come together, it was the physical, crumbly reality of the documents in Possession that showed me how I could do it. In a novel that's all about photography, voyeurism, reflections, and light making images, it's even better to have Anna handling photocopies, with the back of the letter showing through that thin, shiny 1976 Xerox paper, to be read in a mirror: what I'd feared would make the novel more ordinary actually provided the ram in the thicket. And the influence of Possession continues, in a contradictory way. A Secret Alchemy begins thus: "What I have known, I shall not set down. My habit is silence, and it is a habit that has served me well. Words set on paper are dangerous."
And, more broadly, Possession made me think very hard about narrative and voice. When you start a novel, one of the first things you have to decide is, Who's telling this story? Is it me, or an authorial persona not altogether unlike me? Is it some Supreme Being, modern style, who neither comments nor interferes, but merely states what happened? Is it one of the characters now, as the story unfolds or even explodes? Is it one of the characters remembering the past? Or is it really the reader, in which case my black marks on the page aren't definitive but merely the material for readers to make the novel for themselves?
There's no right or single answer, and it makes me very cross when I hear teachers and editors and writers themselves saying that one kind of narration is 'old fashioned' or another 'pretentious'. What Possession showed me is that there needn't even be a single answer for a single novel. As so often in Byatt's fiction it doesn't have to be 'either-or', it can be 'both-and'. 'Both-and' is more difficult to get right, of course. But when did difficult mean that a writer shouldn't try?
A version of this article originally appeared at NORMBLOG