I read a lot of historical novels both as a child and as a young adult. I still have my copies of books by Rosemary Sutcliff, Geoffrey Trease, Ronald Welch and others. I also devoured Jean Plaidyís historical romances (and once heard a fellow medievalist admit in a paper on Edward IV and masculinity that sheíd been heavily influenced by reading her at an impressionable age). Yet now in my limited fiction reading I deliberately shy away from historical novels and find many that I do glance at unreadable.

Before I talk more about my changing relationship to historical fiction I should make two things clear. One is that I am a failed historical novelist. I originally took up researching the early Middle Ages so I could write Charlemagne the Novel. (Actually it was to be called ĎOur Emperor Charlemagneí, after a line in Dorothy L. Sayersí translation of the Song of Roland). After a while I realised that I enjoyed the researching more than the novel writing, and I was rather better at it, and so managed to sneak my way onto a MPhil course. (The rest is research history). So if I criticise historical novels, it is not with any sense of superiority or illusion that theyíre easy to write. Secondly, Iím talking largely about novels set in a realistic pre-nineteenth century past. There have been some wonderful books which have been deliberately anachronistic about historical periods (two personal favourites are No Bed for Bacon and Pyrates. And novels set in a relatively recent past (roughly the last 200 years) not only have the benefit of far more sources to draw on, but also have a subtly different relationship to these sources, because there are contemporary novels from their period. Many historical novels set in the Victorian era, for example, are consciously writing stories that Victorian novelists did not or could not write. I can still read such Ďmoderní historical novels with enjoyment.

Most serious historical fiction set in an earlier period, however, now makes me uncomfortable, and has done so increasingly as my historical sense has increased. The problem for me is anachronism, but itís not predominantly the simple errors of ancient Romans having clocks. The two things I find suspension of disbelief killers are anachronisms in dialogue and social attitudes. (I therefore do not even try to watch historical movies, since I know itíll be painful).

I think I gained my sensitivity to dialogue from writing it myself and learning in that way what sounded authentic while still being readable. I also learned that while itís hard enough to write contemporary dialogue, itís substantially harder to write historical speech and gets more difficult the further back in time you go. Thatís a combination of individual words and phrases becoming anachronistic (I had to think hard about whether I could use a term such as Ďautomaticí in a largely pre-mechanised age), and a lack of sources for historical registers of speech. Before Shakespeare (and to a lesser extent Chaucer) we have very little sense of how a peasant might speak as opposed to a noble, or how a child would address their parents. (I once abandoned a novel on Arthurian Britain at an early stage, because the young Guinevere referred to her Ďmamaí, and I couldnít get past the twee eighteenth century image this produced). Such informal speech (how medieval people talked to their servants, how they swore, what jokes they told) is vital to a sense of characters as real people, but not easily gained via texts such as Beowulf. One of the few things more painful than reading medieval characters say: ĎYou mustnít let this get you downí, is having them say: ĎZounds, my lord, tis twelve of the clock, I must away betimesí.

Even if a novelist can find a Ďneutralí tone for writing dialogue, however, itís still hard for them to avoid their characters having anachronistic social attitudes. While I once read and enjoyed the historical works of Nigel Tranter, for example, I now find his moderate, reasonable, enlightened young men stick out like a sore thumb in medieval Scotland. Oddly, I think that portraying attitudes accurately is something thatís become harder for authors recently, since the 1960s. It may have been easier for female authors raised before feminism to portray naturally a world in which female subordination was taken for granted, just as one of the most authentic portraits of Victorian imperialism was written by George MacDonald Fraser, who clearly shared many of the periodís reactionary views. In contrast, political correctness, which I approve of as a modern phenomenon, is one of the great curses of the historical novel. Most people in the past had views about women, the poor and other races that weíd find offensive today: a good novelist must find a way round this, rather than turning his or her heroes and heroines into liberals before their time.

Anachronism in mental attitudes is also why Iím particularly wary of the historical detective novel as a genre, even though I enjoy detective stories. Detection (or at least its explanation) seems to me to require a style of practical, logical thought that sources rarely show for most periods. I can imagine a medieval man, for example, being able to know that someone was not killed with the weapon suspected, but not explaining in the logical way that a detective plot requires how he knows that. And the detective story as a concept does not really make sense in the 95%+ of historical societies in which it does not actually matter whether the wrong person is found guilty as long as it's someone unimportant.

Professionally as well, I now find even the best of historical novels problematic, the ones that can manage the immense task of feeling right. At one level, itís the basic issue that theyíre not true (or at least not true in some key moments), and as historians we are trained to be wary about the danger of making things up. But perhaps a more subtle difficulty is that historical novels tend to make the past more coherent than it really is. (This is a particular issue for people like myself who know historical novels can be an effective gateway drug to hardcore historical study). Historical novels (and even more novels set in prehistoric times) tell us things we can never learn from the sources: they awake in us a desire for knowledge that we cannot ever obtain: did Elizabeth I have a sex life? What did the people of Skara Brae think about? Most historical novels also have consistent characters and plots: they make sense in some deep way. In contrast, our fragmented sources and the real-life inconsistencies of historical figures often donít form a good narrative. The very skill of successful historical novelists can make the history on which they depend seem thin and unsatisfying in contrast. Historians do not mostly aim to own their readers in the way a novelist might: nor, I think, should they aspire to. But is evidence and a strictly controlled imagination (I now have a whole professional vocabulary of qualifying how secure I think my conclusions are) enough to please those who have experienced the visceral thrill of being sucked into a historical novel?

There are a few historical novelists, who I discovered before I became a historian, who I can still read happily today: Georgette Heyer and Dorothy Dunnett, for example. Maybe there are some others who I ought to discover, but I suspect that my eye and ear has now become too critical to enjoy the vast majority of historical novels. It is a loss to me of one former pleasure, but still I would, on the whole, rather have the Charlemagne of the sources than any Charlemagne of my own fictional construction.