Glenn Burgess On Revisionism
‘On Revisionism’ is an important article from 1990 (you can download it free from Historical Journal) in which Glenn Burgess sets out a fair appraisal of what revisionism is (or was) and defends it from some unfair criticisms, then makes some more sophisticated criticisms. It’s possibly unfair to beat Burgess with a stick that hadn’t been published at the time he was writing and which was (and perhaps still is) considered extremely radical, but I’d like to compare this article with the work of Keith Jenkins, and some of the theorists who inform his work, because there are some surprising similarities. Ultimately Burgess and Jenkins draw different conclusions, but they are tackling some of the same problems and at times use similar arguments. For the purposes of this post I’m not going to question empirical epistemological foundations at all, but if we accept that the past really happened, and that we can know facts about what really happened, there are still many practical and theoretical problems concerning what to do with those facts.
Burgess starts by pointing out that revisionists in the 1970s saw a need to study politics on its own terms because they recognized that politics could not be reduced to anything else. He also points out that even Christopher Hill accepted that the cultural superstructure could not be entirely reduced to the economic base. Brian Manning seems to implicitly support this despite his attempts to defend the base/superstructure model.
I find Herbert Butterfield’s 1931 critique of Whig history interesting because he seems so close to Lyotard, and yet so far away. Butterfield recognized that the entirety of what really happened in the past was too vast too fit into a single book, and that therefore writing a general history involves a great deal of omission. He was critical of the Whigs because he believed that their decisions about what to include and what to omit distorted the truth. The problem which Burgess points out with this view, and where it diverges from Lyotard, is that Butterfield believed that it was possible to write an abridged history which didn’t distort the truth and maintained the one true meaning of history. I agree that this is not possible, and that it invalidates the particular argument which Butterfield tried to make, but he was still onto something with his suspicion that there was a problem with the way Whigs constructed their narrative. Lyotard’s problem with Whig and Marxist grand narratives was not that they distort the truth, but that they present only one truth and claim that this is the only truth. Therefore they suffer from exactly the same fallacy as Butterfield. Arbitrary selection and omission of facts is unavoidable when writing historical narratives, but what can and should be avoided is closing down other possibilities by claiming to have written the only possible narrative. As Burgess points out later in the article, there are many different ways of writing the story of the English Civil War.
Perhaps the most well-known point that Burgess makes is about teleology. He makes a crucial distinction between strong and weak teleology. Strong teleology, which claims that the outcome of a narrative was always inevitable, is obviously wrong. This was a weakness of Whig narratives of progress towards democracy, and Marxist narratives of class struggle. However, Burgess criticizes revisionists for conflating both kinds teleology and trying to get rid of the whole thing. This is an impossible and counterproductive task. Not only is weak teleology not a problem, Burgess shows that it’s absolutely necessary for writing any kind of history. In order to write a historical narrative you have to give it some kind of structure. You have to know where the story is going to end up in order to decide which facts to include and which to omit. There is no objective way of deciding which facts are important because the importance of facts is relative to the story you want to tell. Again this isn’t so very different from Lyotard or Hayden White.
The concept of anachronism is also related to teleology. As Burgess says, strong teleology is anachronistic, which it obviously is. However his definition of anachronism also includes “the use of present day categories to organize our accounts of the past”, and this is described as “reprehensible”. I’m not convinced by this part of the argument as I’m not sure that it’s possible to write about the past without using any present day concepts. Hayden White, Frank Ankersmit, and Keith Jenkins would all say that everything in history above the level of very basic facts is metaphorical. In what way can concepts in present day minds not be present day concepts? Maybe it would be possible to write a historical narrative which uses only contemporary language and concepts but that would be a pastiche of contemporary writing. That would be an interesting possibility, and I hear that Ariel Hessayon has attempted something like it, but I don’t think it’s what Glenn Burgess was calling for. In fact his critique of Clarendon implies that he would not consider such an approach to be “proper” history.
Burgess insists that Clarendon “is not in any modern sense an historian at all” because “his purposes are consequently (in our terms) moral rather than historical” (p. 622). Keith Jenkins deconstructs the concept of “bias” showing how the dismissal of some sources and historians as biased allows the historians doing the dismissing to believe that they are not biased themselves and have achieved objectivity. Burgess is clearly falling into this trap by identifying Clarendon as the Other who is not-history, against whom “proper” historians can define themselves. It’s not clear to me that having a moral purpose stops writing from being history (I dare anyone to try telling me that feminist history isn’t proper history), or that it’s possible to write without a moral purpose. Even trying to exclude other people’s morals from history could be seen as a moral purpose of sorts.
His other problem with Clarendon is that “human wickedness and incompetence, without providence will produce only chaos. And chaos has no story, no history” (p. 623). First of all I think that history could well be chaos in the scientific sense of chaos theory, but I don’t think that this is what Burgess was getting at. He probably meant it in the more common sense of randomness or lack of order. But perhaps history is random and disordered, in that the body of knowable facts about the past is so vast and unintelligible that it might as well be random until we impose some kind of order on it by structuring it into a coherent narrative (with all the arbitrary decisions which that entails). It is entirely possible to construct a narrative which tells a story of random things which don’t happen for any apparent good reason. Burgess more or less concedes this by characterizing Clarendon’s work in pretty much these terms (and for even more extreme examples of narratives of chaos see Samuel Beckett or Franz Kafka). For Burgess Clarendon fails to be proper historian again because he fails to explain things properly, but this is all down to a difference of opinion between Burgess and Clarendon about what counts as a proper explanation.
This is not to say that Burgess has a narrow view of what historical explanation should be. He’s against determinism, strong teleology, and randomness but he acknowledges that many different explanations are possible and that choosing between them might not be possible. As an example, he takes a model of the civil war as baronial revolt, based on Conrad Russell and John Adamson, and contrasts it with a model of the civil war as war of religion, based on John Morrill. He points out that these two explanations are not necessarily mutually exclusive and could just be different ways of constructing a narrative about the same events. Both might be true but incomplete, describing different parts of the same problem. This suggests that many more narratives of the same events might be possible. I suspect that, as Michael Braddick has recently suggested, the story of the English Civil War will never be complete because there will always be more aspects of the story which could be told, or aspects which could be told differently. Historiography has traditionally been much more adversarial than this, consisting of knocking down old interpretations and replacing them with something else, which in its turn gets knocked down and replaced. Perhaps revisionism was the beginning of the end of that process. Whig and Marxist narratives clearly needed to be cleared away because they claimed to explain everything and denied opportunities for different views. If revisionism hasn’t put anything similar in their place that could be seen as a success rather than a failure (although there are many ways in which revisionists have tried to close down potentially valid possibilities, such as Conrad Russell’s refusal to engage with the question of revolution).
There are still traces of the adversarial approach in the article, as Burgess went on to look for ways of choosing between explanations. He points out that explanations of the same thing might not be about the same thing if you look closer. The relative strengths and weaknesses of an explanation depend on what it’s trying to explain. If you choose a different end point you get a different narrative with a different explanation. Therefore how you explain the civil war depends heavily on how you define it. This is all good. At this point it seems to me that the logical conclusion should be that because different people choose to write different narratives with different end points their explanations are always going to be different and that there might not be any way of choosing between them (or any need to choose between them). And above all there’s no objective way of deciding what you should try to explain, or deciding how to define the problem. It’s an arbitrary decision which depends on things like personal taste, audience expectations, and yes, moral purpose.
Surprisingly, the final section doesn’t go where the rest of the article seems to be pointing. Burgess ends up trying to find an objective way of choosing between explanations by looking for an objective way of choosing between end points. Against all the odds, he seems to think he’s found it. He concludes that what we should try to explain is what most needs explaining, and that what most needs explaining is what was most unusual or unique. I think there are lots of problems with this. First there’s the problem of defining the problem. Burgess earlier argued that historians find different explanations of the Civil War because they have different definitions of what the Civil War was (and up til now he implicitly accepts that there is no objective way of choosing between these definitions). At the beginning of section V he suddenly jumps from using the term “Civil War” to using “the English revolution”. This is highly significant, because from this point on he maintains (without explaining exactly why) that we can objectively know exactly what the English revolution was, and can objectively compare it with other historical events to see what the similarities and differences were. I don’t see how this follows at all. If we can’t define the civil war, how can we define the revolution? Don’t the same problems apply to both concepts, or neither? To me “revolution” looks like an abstract/analytic concept which can be (and in practice has been) defined in many different ways. There can’t be an empirical test for a revolution without a definition to test against. If you don’t define a revolution you can’t find one, which is possibly why Conrad Russell was so reluctant to even discuss the question.
Even if we could agree on what the English revolution was and determine which bits of it were most unusual (tricky, since Burgess himself says earlier in the article that things which appear the same might not be the same when you look closer), would this mean that we had to focus only on the unusual things and ignore what was usual? I disagree for three reasons.
First, because we don’t need to limit ourselves to writing about things which are generally agreed to be important. There might be many interesting works to be written about “usual” things, and there are more than enough historians to go around. We don’t all have to write about the same thing.
Second, because this line of argument seems to assume that some things, such as baronial coups, are just conditions of a certain period. Steve Rigby convincingly argued (although a few years after this article) that causes and conditions can’t be distinguished objectively, because it depends on what historians and their audiences are prepared to take for granted, which might vary a lot. Although Burgess means a slightly different thing by conditions I think Rigby’s argument is still applicable here, because it ultimately concerns what to take for granted and what to explain. Some people might not think that baronial coups need explaining because they were just endemic in pre-modern England, but other people might want to know why they kept happening, and why they didn’t stop earlier, or carry on later. When Burgess says “There is nothing particularly unusual about either baronial revolts or religious wars” he seems to be saying the same thing that he earlier criticized Peter Laslett for saying.
And third, because accepting something as “just how it was” makes it appear natural and puts it beyond criticism. Unequal distribution of resources and oppression of women are basic facts of history, and not particularly unusual. At what point do they stop being “just how it was” and start being contemporary problems that we have to deal with in our own lives? (The discussion over at Smashing the Window about the tension between history and contemporary history is particularly relevant here.)
Every historical narrative necessarily has to take many things for granted, but I don’t think it’s healthy if we all take the same things for granted.
Finally, one of the biggest problems of causation is hinted at, but not dealt with in any detail. Burgess rightly points out that when Conrad Russell found the outbreak of civil war in 1642 to be a peculiarity of Charles I’s reign which was unlikely to have happened to Henry VIII, he needed to explain what was different. “The account is incomplete, dependent on something else which is itself in need of explanation” (p. 623). Or, as Keith Jenkins asked “how far back and how far afield” do you need to go? Only you can decide, and your decision will be arbitrary.
Overall this is a fascinating but frustrating article as it seems to get so close to a postmodern breakthrough but then pulls back at the last minute. It could have had a more radical conclusion, but at the time it was written that wouldn’t have been fashionable and would have aroused a lot of hostility, so perhaps it isn’t fair to criticize Glenn Burgess for not being too far ahead of his time. I wonder how far he realized the implications of what he was nearly saying, and how he would approach these questions today, now that Michael Braddick and Malcolm Wanklyn are quite happy to accept some postmodern ideas.
- Glenn Burgess, ‘On revisionism : an analysis of early Stuart historiography in the 1970s and 1980s’, Historical Journal, 33 (1990), pp. 609-27.
- Keith Jenkins, Re-figuring History (Routledge, 2002).
- Keith Jenkins, Re-thinking History (Routledge, 2003, first pub. 1991).
