I have always tried hard not to mysticise the creative process. To me, imagination, although precious and powerful, is not supernatural; ideas come to us through a process of conscious and subconscious suggestion. Writers draw upon facts, or experience; they write from a position of knowledge or in pursuit of it. There’s no magic, sadly. No divine intervention.

Or so I thought. If, one month ago, someone had told me that a writer of historical fiction could utilise the ‘hunch’, that inexplicable, intuitive persuasion that comes out of the blue, uncalled for, unfounded and seemingly irrational, to achieve historical accuracy and verisimilitude otherwise only attainable through research, I would have sneered. However, that was one month ago.

In August this year I travelled to Iceland to research the life of a woman called Agnes Magnusdottir, a servant born in 1795, and the subject of my PhD. Agnes Magnusdottir was the last person to be formally executed in Iceland before the death penalty was abolished in 1928. Beheaded by broad axe in 1830, at the age of 34, Agnes was killed for her role in the violent murder of two men and the act of arson that attempted to burn their bodies.

I first learnt of Agnes in 2003, but it wasn’t until 2007 that, motivated by a frustration at the prejudice and contradictions evident in accounts of the murder and execution, I decided to write her story. My novel would attempt to represent Agnes as more ambiguous than an ‘inhumane witch, stirring up murder’, and also address the inconsistencies of the historical record. In order to write it, I decided, I would need to undertake rigorous primary research. While I understood that imagination would be necessary – the book was, after all, a novel – everything, I decided, would be as factual as possible.

After one year of research I realised that, despite now having a considerable amount of information about the murders, I had, miserably, absolutely no biographical information about Agnes whatsoever. How on earth could I begin to write her story with historical accuracy without even a date of birth? To counter the total lack of sources, I decided to invent a fictional life for Agnes through a process of ‘logical speculation’. I researched early nineteenth century Icelandic life obsessively until my dinner-party conversation reached a new low (‘Did you know that stored urine forms ammonia? Icelanders would clean their hair with it. Ooh, is that the sav blanc?’), and used juicy snippets of historical trivia to trigger biographical speculation. A scholarly article on high infant mortality in nineteenth-century Scandinavia led to my supposition that Agnes would probably have had a sibling who died. An 1816 diary describing the prevalence and nature of poverty in Iceland led to my decision that Agnes would have been similarly destitute. Maps, journals, academic articles, song lyrics, photographs, and even tourist brochures began to feed into my fictional construction of Agnes’s early life.

However, despite my pure, scholarly intentions, I soon realised that not all of my draft was based on ‘fact’.  I became increasingly aware that some creative decisions were being influenced by what I can only describe as ’hunches’.

This realisation came a few thousand words into the first draft, when, pausing to mop up coffee I had spilt all over my keyboard, I felt, intuitively, that Agnes had grown up without a mother. It was a weird, unprompted certainty, and, as far as I could tell, totally unfounded. I wondered if it was some symptom of excess caffeine consumption (the spilt cup was my fourth of the day), but the retrospective premonition was impossible to ignore. Agnes grew up without a mother. I knew it. I tried to rationalise: perhaps this hunch was from a subconscious belief that crime originates in childhood trauma; or from a desire to make my protagonist sympathetic; or from a suppressed and rather base longing to make my novel as dramatic as possible. I had no idea.

Soon after I had this hunch, I became aware of other intuitions. For example, I felt that Agnes had a younger brother. Again, I didn’t know if this gut feeling was the result of subconscious, applied logic, or something a little more mysterious, but when I began writing an episode from Agnes’s childhood, I gave her a three-year-old brother.

My draft started to take shape. For the most part it was composed out of logical speculation and yet it also included events and characters that had no basis in my research: they were singular, intuitive guesses. Agnes and her brother were separated. Agnes did not have any children. The priest who was with Agnes in the last months of her life was inexperienced. Some of these hunches came from dreams. Others came out of the blue. For the most part I went with them, reminding myself that the novel was ultimately a work of fiction, and that the certainty I felt in my intuition most probably came from my commitment to the creative process, rather than from some mystical enlightenment.

Then, with 40,000 words of my novel down, I was awarded funding to travel to Iceland for the sole purpose of continuing my research into Agnes’s life. The news, while exciting, was also nerve-wracking. I had promised myself, and my supervisors, that I would stick to the facts. What would I now find out? Would I have to completely re-write my account of Agnes’s childhood? I kicked myself for caving into what I thought were the whims of my imagination.

On my arrival in Iceland, I immediately began to visit the National Archives, the Arni Magnusson Institute, and other libraries around the country. Having very little personal documentary evidence to point me in the right direction, I began by poring through the national census of 1816, looking for any listing of Agnes Magnusdottir in the area of the execution site. I found her after only a morning’s work: Agnes Magnusdottir, a 21-year-old servant working at Guðrúnarstaðir, born in Flaga. It was then I was able to look for the record of her birth in the ministerial books of the parish in which Flaga stood. After several hours of skimming through lists of spidery cursive, I found it: Agnes Magnusdottir, born 27 October 1795 to Magnus Magnusson and Ingveldur Rafnsdottir, unmarried servants. With these names, places and dates in hand, I continued researching tangentially over the next two weeks; flipping through census data, consulting ‘soul registers’, comparing ministerial records and tracking family members as they entered and left different parishes in the district. As with all primary data, the facts revealed themselves slowly, but surely in my notes. Soon my skin was crawling: all of my hunches, all of them, were more or less on the mark.

I was astonished at first, then disbelieving. I checked the records again, I re-read my notes; it was uncanny. Not only was Agnes’s childhood similar to what I had speculated in light of my research into nineteenth-century Icelandic life (Agnes did have a sibling who died young: a half-sister, Helga, who passed away at twelve), but it was also resembled, even more closely,  what I had intuitively felt was the truth:  Agnes did have a brother, a half-brother, five years younger (exactly the same age difference I had intuited); Agnes’s mother was absent from her childhood – she evidently abandoned her daughter when Agnes was only six; Agnes was separated from her brother – her mother took him with her, and then also abandoned him, at a later date; the priest was inexperienced – he was listed as an assistant pastor.

It was a thrilling, albeit eerie series of realisations.  As someone who, as I said before, tries not to imbue or associate the writing process with the mystical (‘Writers are sensitive geniuses who require bohemian dress, mouldy attics and pneumonia to lure the muse’, etc), to have had inexplicable creative premonitions confirmed by the historical data made me wonder whether I had been a little too rational for my own good. Was someone, or something, helping me out?

Of course, hunches have been around for a long time; they occur too often to be put down to sheer coincidence. In today’s scientific, hyper-rational world, neuroscientists attribute the ‘hunch’, or seemingly illogical intuitive knowledge, to the medial frontal cortex of the brain. Largely an unconscious action, they claim that intuition is what happens when your brain collects information from ‘facial expressions, body language, emotions, familiar and unfamiliar patterns, oddities and routines, things visible and invisible, inconspicuous and unremarkable’, and then filters and interprets it ‘at a speed infinitely greater than in any other region of the brain’.

This certainly makes more sense than intuition as supernatural power. My hunches arrived after a considerable amount of research, and it is certainly possible that my general knowledge of nineteenth-century Iceland buzzed around in my medial frontal cortex for a while, and then shot through some ideas as if out of the blue. But how does that explain the very personal nature of these hunches, their striking specificity? The five-year age gap between her and her brother? The absence of her mother? It doesn’t seem possible. Is it a combination of research and imagination? Intuition and coincidence? None of the above?

In the meantime, my early intuitive narrative decisions have saved me months of re-writing, and for that I’m grateful, if not a little spooked. I’m continuing my research just the same, but now, each time I feel a tingling certainty, seemingly without foundation, I take it more seriously than before. My correct hunches about Agnes might be thanks to some secretive, whirring part of my brain, or a mystical external force – I’ll probably never know – but either way, they feel pretty magical to me.