My eureka moment as a historian came on a snowy afternoon in January 1977, in the Upper Reading Room of the Bodleian Library, Oxford, when I was a week into my second term of postgraduate study at the university. The subject I had chosen for a thesis was the Royalist war effort in the English Civil War. To put it into a manageable framework meant settling for a regional study, and the areas of the country that seemed to have the best surviving records for the subject were Wales, the Welsh Marches and the West Midlands. Choosing those areas had two other advantages for me - one practical and one emotional. The former was that most of my destinations could be reached fairly easily from Oxford, by hitch-hiking or train, the two standard means of travel for a student in the 1970s. The latter was that much of my original interest in the Civil War had been kindled in my early teens during summer holidays in Wales and its border counties. It is a land in which the bones of history lie around upon its surface, having been not as commonly recycled for building material or ploughed away by agriculture as in the southeastern English county in which I had grown up.

In particular, I was attracted by the ruined castles that littered it, and as I visited these, and bought the guidebook to each, I had become aware of the importance of the great struggle between Charles I and Parliament in giving the castles some of their most dramatic history, and turning them into the shells and wrecks that most now were. The roots of my postgraduate work lay in that process of adolescent discovery. So there I was in the Upper Reading Room of one of the most beautiful libraries in the world, whose golden fluted limestone walls had made my heart soar when I first saw them the previous summer. Surrounding me were portraits of worthies from the period I was studying, and ranked on the shelves behind were the huge, heavy, leather-bound tomes that contained some of my most important source material. Like most hard-working research students, I had devoted the autumn term to reading through all the secondary sources for my subject, the works published on it by historians. Now, equally on schedule for a keen postgraduate, I was starting the process of reading the primary texts; the moment at which the actual business of research can be said to begin, and equivalent to the experience of a soldier who, after months of training at base camp, finally goes into action for the first time.

One of the most important secondary sources for my project was Memoirs of the Civil War in Wales and the Marches, J.R. Phillips' two-volume narrative history published in 1874. It was ponderously thorough and meticulous, and had a large appendix of primary sources. As a result, not surprisingly, previous historians of the subject had treated it as authoritative. One of the more exciting passages in the book described two successive major battles fought between the Earl of Stamford, the Parliamentarian commander based at Hereford, and the Marquis of Hertford, the Royalist general in southeast Wales, in November 1642. This was the first winter of the war, and the two actions represented the first major military clash in the area. In both, the Royalists had been crushingly defeated. The first had been precipitated by an attempt on Hertford's part to bypass Parliamentarian Hereford as too strong to take, and conquer northern Gloucestershire instead. Stamford had chased him and brought him to battle near Tewkesbury on 15 November. The Royalists had 7,000 men against 4,000 Parliamentarians, and Stamford's infantry consisted of a single regiment of soldiers reinforced by local militiamen. Nonetheless, the Parliamentarians had better training, muskets and artillery, and were a more cohesive and motivated force. As a result, they stood firm when the Royalist horsemen and then their Welsh foot soldiers attacked. It was their assailants who broke and fled, and suffered fearful losses when pursued by Stamford's men: 1,200 taken prisoner and 2,500 killed, including a prominent gentleman, Sir Rhys ap Huw Granock.

An action on that scale counts as one of the larger battles of the whole war. Only 12 days later, however, Hertford had been reinforced and now advanced directly on Hereford, determined to avenge his defeat by attacking when Stamford had sent away many of his men and had only 1,500 in the city. The Earl came out to meet his enemies, holding them up with concentrated fire until thousands of local people rallied to reinforce him. The Royalists then retired, only to be ambushed by a party that Stamford had hidden on the line of their expected retreat, and put to flight. They lost 2,000 more men, and Hertford himself escaped only by hiding in a wood. With that, the Royalist threat to the Wye and Severn basins was ended, at least for a season.

ÖйúAƬ

As was his habit, Phillips reprinted the source material for this narrative in his appendix, consisting of two pamphlets published in London that respectively provided the stories of the two battles. Stylistically they are similar enough to have come from the same pen, and each was printed within a few days of the events described, apparently when the news first reached the capital; which would account for the vivid detail of the descriptions. They are confident, compelling primary sources of the classic kind. That is why it is so significant that I was in the Upper Reading Room that afternoon, and that the huge volumes behind me included the journals of the House of Lords, one of the first bodies of evidence I tucked into, outside the work of historians. They mattered because they included printed copies of dispatches from Parliamentary commanders, and one of these was the Earl of Stamford, writing from his base at Hereford in November 1642. He boasted of the single action that he had actually fought: a raid on a Royalist encampment that killed 21 of the enemy. Otherwise he complained endlessly of the hostility of the local people and his lack of money and supplies. As November gave way to December, the enemy outposts crept ever closer, and his nerve gave way, so that he abandoned city and county to his opponents, and retreated out of the region.

I realised at that moment that the battles had never happened: the pamphlets had been packs of lies, deliberately concocted to cover up a true story of total failure in the region. More important, the whole war was going badly for Parliament at that time, with the King's main army advancing on London. Somebody had tried to bolster morale by inventing tales of major victories in a theatre so remote that nobody in the capital was likely to check their accuracy. Whether or not they had actually fooled their intended audience, they had deceived historians for more than 100 years; indeed, because of the lag in the publication and reception of my work, they were still being repeated as fact in books that appeared into the 1980s.

ÖйúAƬ

What I felt in that moment, in my own miniature way, was that godlike anger that had fuelled events such as the Reformation and Enlightenment: the realisation that what I had tamely accepted as fact, from the mouths of my elders and betters, had been a fraud. Perhaps more to the point, I bought into, for the first time, that suspicion and disrespect towards established authority that had been a hallmark of many of my generation in the previous decade. Those rebels had warned us not to trust anybody over the age of 30: I suddenly wondered if you could trust any historian over the age of 25.

To put this in perspective, it is necessary to appreciate that I had been educated, at school and (substantially) at university, to view the progress of knowledge in the Victorian manner, as a cumulative process by which each generation filled out more of a story handed on by its predecessors. Exponents of this tradition looked back to the pioneering scholars of the 19th century as their exemplars, just as Britain's economic, military, cultural and social attitudes were articulated during my childhood in terms formed 100 years before. As youngsters we had been given the novels of Dickens, Rider Haggard and Jerome K. Jerome to read as if they were still fresh and relevant. Suddenly all that crashed in on me, once and for all, and I turned into a revisionist. I also, however, bought into, even more strongly than before, the equally old-fashioned idea that ever-better historical scholarship would bring us closer and closer to the truth of the past: I had just exposed a lie and established a historical reality, and done so in my second week of hands-on research. Throughout my work on the thesis I was to have that exhilarating sensation of putting one piece after another of a jigsaw into place, making a picture never seen before.

I was in advance of my immediate peer group in this realisation: when I went from the Bodleian to a dinner party of fellow students that evening, my friends were incredulous at what I had found. In another sense, however, I was wholly of my time, for this was the period in which Britain was letting go of its Victorian past: losing the Empire and great-power status, preparing to bid farewell to an economy based on heavy industry and a social system based on class, and shedding a network of restrictive cultural attitudes. Just as it was having to replace a physical infrastructure of 100-year-old sewers beneath cities, so an intellectual infrastructure was being supplanted. I made my personal discovery just as radical revisionism was emerging as a great force in Stuart historiography. A few years later, one of its leaders, Conrad Russell, told me that he had experienced the same epiphany, in the earlier part of the decade: he felt that he could believe that there had been kings called James I and Charles I, but practically nothing else that an established historian told him.

Of course, I was overreacting. The elimination of those two battles from the record did not change anything significant in our view of the Civil War. I found that I was not even the first person to identify them as fraudulent, the job having been done by a local historian of Herefordshire, John Webb, in the mid-19th century (although nobody had noticed this). Nor was my sense of a progressive achievement of "true" history capable of standing up well to the postmodern challenge of the 1990s; I had certainly found a truer one, but ultimate accuracy would always be unattainable, and the "why" in history would always be much harder to prove than the "what".

ÖйúAƬ

Nonetheless, I had immediately been given an assurance that, even at my junior level, history was worth writing. I had also been given a taste for blood. I have never savagely attacked the work of another living scholar: personal combat is not natural to me. But a suspicion of historical icons has remained with me ever since: within three years I was to question the reputation of The History of the Rebellion and Civil Wars in England: Begun in the Year 1641 by the Earl of Clarendon, traditionally the most admired historian of the Civil War, and I have left a trail of such targets in my wake since, the most recent being the Iron Age bog body in the British Museum, Lindow Man, which had been interpreted as solid evidence for human sacrifice. In an important sense, I have never looked back from that moment in the Bodleian Library, 33 years ago.

Ronald Hutton is professor of history, University of Bristol, and author of Charles the Second: King of England, Scotland and Ireland (1989) and The Royalist War Effort: 1642-1646 (2002).