What historical fiction can't teach you
Happy new year, readers! Hope it's a cracking one for all of you.
Since it's the season for curling up by the fire with a weighty tome and a whisky, I thought I'd kick off 2018 on the blog with some reflections on what is - to my mind - a peculiar, even apparently paradoxical, fact about historical fiction, which I don't think has been much noticed before. Namely: historical fiction - by which I mean, fiction heavily influenced by knowledge or beliefs about the past - is apparently less reliable as a source of knowledge than fiction which is not explicitly historical. If that's right, then it's surely surprising.
My worry isn't about shoddily researched historical fiction, where the material drawn upon is itself unreliable; or where details are added without care for accuracy, but nonetheless are presented misleadingly as authentic. There's no surprise in thinking we can't learn much from that. Instead, I have in mind fictional works by authors whose research practices meet the highest academic standards. Think of Wolf Hall by Hilary Mantel, about Thomas Cromwell, a key player in the court of Henry VIII. Mantel spent five years researching the book, and aimed only to include invented detail where it was compatible with established fact about Cromwell and those around him. Or think of Regeneration by Pat Barker, which brilliantly incorporates in a fictional narrative many real-life facts about the stay of both Wilfred Owen and Siegfried Sassoon at Craiglockhart Hospital in Edinburgh in 1917, under the care of the psychiatrist W.H.R. Rivers.
Now, at the time of reading these books, I certainly got the impression I was learning things. How could I not be, when both authors had been so careful to infuse the scenes they created with authentic and meticulously researched historical detail? But since then, I've thought about it some more, and concluded that even if I did gain many true beliefs from those works (about Cromwell's life in the court; about Sassoon's life in the trenches, and afterwards), there's good reason to think I couldn't have got knowledge.
To see this, we need to take a step back, and first examine a worry people have had about fiction which contains true statements, generally (that is, not a worry about historical fiction per se, but about any fiction which contains true information, even in passing). Fiction often does this: for instance, if a novel is set in London, then, alongside lots of invented details, it might contain some true details about London streets. These true statements might be included to give a sense of 'authenticity', or 'realism'; or maybe the author was lazy and just couldn't be bothered to make that stuff up. The reason is irrelevant for current purposes.
The worry can be expressed in terms of an analogy (by way of a thought experiment). Perhaps fiction with true details scattered here and there is a bit like a landscape in which there are lots of fake barns, and you are driving about, trying to find a real barn. Say that, against the odds, you come across a real barn. You might just as easily have come across a fake one, indiscernible from a real one - after all, in this country, there are a lot of them about. Say that now, as you stand in front of a real barn, you have the true belief 'there is a real barn in front of me'. Even though this belief is true, do you know that there is a real barn in front of you? Some philosophers have thought not: for the presence in your vicinity of many fake barns, and the closeness of the possibility that you might have been standing in front of one of those, and not a real one, is enough to undermine your current true belief's claim to knowledge.
Analogously, one might worry, even if you happen to get a true belief from an accurate description in a fiction, the presence of lots of invented false details surrounding that description make it the case that you might just have easily believed something false; in which case, your current true belief doesn't get to count as knowledge.
On reflection, there's something wrong with this analogy, however. In fake barn country, even when I am in front of a real barn, I can't discern any difference between it and a fake one (and vice versa). That's built into the thought experiment as a given. However, when it comes to fiction, in many cases, competent readers can discern true details when they come across them in the midst of a largely invented scene, and distinguish them from false ones. There are a range of ways in which an author can intentionally signal to a competent and experienced reader that a particular proposition is to be believed, and not only imagined. At a minimum, she might put it in a declarative sentence and an authoritative tone. She might also include information concerning existents the reader is likely to already know about, and that builds on information she already possesses about them. She might also include information that, she infers, the reader has some interest in knowing. There might even be some convention in place, such that wherever a real existent probably known to the reader is mentioned (a town, a park, even a historical person), it is always described accurately. And so on. In other words, in virtue of the conventional manner in which they are introduced and formally presented, readers can often discern true sentences in fictions from surrounding invented and false ones. And if this is right, the worry as originally introduced disappears.
However, unfortunately, it re-emerges as a worry about historical fiction in particular. For properly researched historical fiction doesn't just have true descriptions occasionally scattered here and there. It is saturated with them. Moreover, there will likely be no easily identifiable formal or conventional difference between the ways in which true and false descriptions are introduced in these works. The very same historical personage may be described accurately at one point, and falsely at another (after all this is historical fiction, not history). Many apparently authoritative declarative statements will be made, and many of them will be invented. The reader won't be able to distinguish those facts she might have an interest in knowing, from those pseudo-facts which are made up, since they concern the very same people and events. And so on.
Of course, I don't deny that someone with the right sort of background knowledge could distinguish the true from the false in Wolf Hall or Regeneration. Historians of the relevant periods presumably could, as of course could the authors themselves. But the average reader is not a historian. So, unlike in the non-historical case, it looks like she really is in the position of the driver in fake barn country. If that's right, then meticulously researched works of historical fiction are worse, as sources of knowledge, than largely invented works which happen to contain true statements in passing. Which is a bit of a shock.
If you're interested in the detail of this issue, I discuss it further in my contribution to this fine new volume, just out with Oxford University Press.