The organizers of the Walter Scott Prize for historical fiction define a historical novel as one that is set at least sixty years in the past. That covers an extraordinarily wide range of work. At one end of the spectrum you have a swashbuckler like Stanley J. Weyman’s Under the Red Robe, which transports the reader (when I was young, I was an eager passenger) to an exotic world of hot-blooded gallants and panting miladies; at the other, novels such as Martin Amis’s The Zone of Interest, where the journey back in time is prompted, not by escapism, but by the urge to confront an historical event so monumental that it still dominates our moral landscape today.

Both these ways of writing about history seem to me perfectly valid. But my own novels – covering, between them, the nearly 200 years between the late eighteenth and mid-twentieth century – spring from a slightly different preoccupation: the feeling that, to understand the world we’re in, we have to engage imaginatively with the world we came from. That means making a deliberate effort to rid our minds of some huge – but largely unconscious – obstacles.

The first and most obvious stumbling block is hindsight: the tendency, when you’re writing about 1788 or 1913, say, to read the characters’ lives in the light of what we know (and they of course didn’t) was about to happen to their world. But there are subtler temptations, too, stemming from our contradictory ideas about the direction of time itself. On the sunny side of the street is the Whig conception of progress, which – since it views history as a one-way journey of constant improvement – effectively condemns earlier generations to a more ignorant and benighted existence than ours. On the other side, lurking in the shadows, is the suspicion that everything’s going to the dogs, and that the past was somehow simpler and more natural than the present.

Both of these tendencies are dangerous because they prevent us from seeing the full complexity of our ancestors’ experience as they wrestled with a reality just as baffling and unpredictable as our own. While all historical fiction is inevitably a product of the time in which it is written, its ultimate value, I think, depends on the autonomy and truthfulness of the voice it is able to give the past.

And the notion of voice is not just a metaphor. Hearing the past, for me, is as important as seeing it. Indeed, the two things are inextricably connected. Language is not simply a dressing-up box of individual words and phrases, which can be draped over an unchanging core of meaning: it incarnates the meaning because it is the apparatus by which we organize and make sense of reality. So it is only when I’ve mastered how my characters think and speak that I feel confident enough to start looking at the world through their eyes — noticing the details they would notice; interpreting what happens to them in light of their assumptions, rather than mine.

Some years ago, when I was explaining to a fellow-writer what I do, she burst out, ‘Oh, you mean you write pastiche!’ It is a common response — but my answer is: no, absolutely not. Pastiche, it seems to me, is a crude form of signalling: it involves the use of well-worn tropes – a phrase, for instance, like Jolly good show! – that have come to evoke a period in the minds of modern readers, but have lost (if they ever had) any organic relationship with the reality they are meant to represent.

An obvious analogy is with architecture: a neo-Georgian house will use a few idioms from the classical phrase-book – pediments, porticos, fanlights – to suggest, to a contemporary eye, a (supposedly) more gracious and tranquil age. The interior – with the possible exception of a fake fireplace or two – will be entirely modern, however. What I do, by contrast, is try to work from the inside out: I learn the (or an) English that’s appropriate to the world in which the novel is set, and then use it – exactly as I would my own present-day English – to describe the characters’ experience as vividly and authentically as I can. In doing so, I’m not pastiching anything, any more than I’m pastiching French when I speak to our friends in Normandy in their own language.

I felt vindicated in this approach when the novelist Allan Massie, in a Scotsman review of my first novel, The Dark Clue, wrote:

...Wilson hasn’t written a pastiche Victorian novel, as, for instance, Charles Palliser did in The Quincunx. Indeed he has brought off something which is really very difficult: to write a novel set in the period when the greatest English novels were written, and to come up with something which is not an imitation but never rings false.

But, of course, the language varies according to the period, which means that every book demands a different strategy. The English required for two of mine – the mid-nineteenth century The Dark Clue and the pre-First World War Consolation – felt so familiar to me, from the Victorian and Edwardian literature I read when I was growing up, that writing them was simply a matter of tuning in.

But developing a voice for The Bastard Boy – set in the lead-up to the American Revolution – posed a greater challenge. By way of preparation, I re-read The Expedition of Humphry Clinker and Tom Jones; but – partly, I think, because of the constraints of the early novel form itself – neither of them seemed quite to have the energy and inventiveness I wanted. So I turned instead to the Letters of Laurence Sterne, which are gloriously ebullient, and full of characteristically wild metaphors and flights of fancy.

But it is not just the vocabulary and the conceits that distinguish the liveliest eighteenth-century English from its modern counterpart: it is the rhythm and structure of the language itself. Sentences often have a spring-like tension, holding two contrasting elements in opposition (a reflection, perhaps, not only of the classical idea of thesis and antithesis, but also of the Enlightenment passion for precise classification). So – after some trial and error – I found myself composing passages such as:

I can remember now few particulars of that meal; for it is in our nature to forget most what we understand least […] But if the matter of the conversation is long gone from me, its manner remains stamped on my memory […].’

Eventually, when I’d had enough practice, writing in that voice became second nature to me. And when it did I found it starting to arrange my thoughts and perceptions, enabling me – if only for a moment – to glimpse an eighteenth-century view of the world.

At the other end of the scale was my novel The Woman in the Picture, which follows the life of a young English filmmaker, Henry Whitaker, between the wars. Here the difficulty lies not in the superabundant richness of the contemporary language but its anaemic flatness – which, indeed, forms part of the subject of the book. The more I researched, the more I realized that the retreat into ‘plain’ English was to some extent an understandable response to the First World War and the florid Victorian rhetoric that had justified it.

This distrust of emotion drove some novelists to seize on the emerging art of cinema – with its cool, supposedly objective gaze – as a model. Writing about Graham Greene, Evelyn Waugh said:

It is as though, out of an infinite length of film, sequences had been cut which, assembled, comprise an experience which is the reader’s alone, without any correspondence to the experience of the protagonist... [T]he cinema... has taught a new habit of narrative.

So how do you find a voice that’s true to this sensibility, but still linguistically vivid? The answer came, again, from writing – such as the journals of the Surrealist poet David Gascoyne – that wasn’t originally intended for publication, and tends consequently to be more relaxed and colourful. So the bulk of The Woman in the Picture consists of a series of excerpts from Henry’s diary – interspersed with passages set in the present, as his daughter Miranda tries to unpick the riddle of his life.

The Summer of Broken Stories, my most recent novel, introduced a new element to the mix that I found slightly unnerving. It is set at the end of the 1950s — right on the cusp between the vanished, more-than-sixty-years ago world of historical fiction, and a just-still-alive memory of my own childhood; so that when I was writing it, I found myself simultaneously in the roles of the archaeologist and of the jumble of bones he’s trying to magic back to life. The result was a collision between the clichéd view of the fifties – which sees it as little more than a smug, monochrome prelude to the psychedelic explosion of the sixties, peopled by prim women and buttoned-up men with moustaches – and the uneasy, shifting-sands reality I found on my return visit to my ten-year-old self.

Replaying in my head the voices I grew up with, I realized that far from being monolithically conformist they were extraordinarily diverse, reflecting differences not only of class and geography, but also, crucially, of generation. In particular, I was struck by how far my parents and their contemporaries were still haunted by the war, and how the ghost of that experience affected what they did (and didn’t) say — while older people, who had grown up before the First World War, tended to be far more forthright and plain-spoken.

It rammed home to me, yet again, that, if we’re to understand who we are, we must listen to the past that produced us in all its complexity, rather than just caricaturing it as either a lost idyll or a world of blinkered ignorance and prejudice.

James Wilson is the author of five novels. The latest, The Summer of Broken Stories, was published in May.