Oh yeah? Says who?
In several previous posts I’ve expressed concern about the pressures encouraging historians to soft-pedal or even minimize the evidence supporting their claims in their published works. In our current age of decreasing attention spans and declining academic books sales, there is growing pressure to eliminate the footnotes (or at least relegate them to annoying endnotes) and shorten the lengths of scholarly tomes by eliminating discussion of historiography and cited evidence. Some historians even counsel us to cull the evidence for our claims to a single example. What’s a historian to do?
Do we really need more than one example?
In part this practical question of how much evidence to pack into your prose fits within a much broader tradition of arguing over which historical questions are most important to ask, and how best to answer them. The tension between narrating and analyzing has always been one of the distinctions between popular and academic history, while generations of academic historians of all stripes (positivist, annaliste, social, cliometric, cultural, narrativist, microhistorical…) have debated the ‘proper’ types of historical questions and evidence, their answers conditioned by the widely-varying amount of evidence they have to draw from. Social and economic questions tend to be answered with quantitative data presented in tables, graphs and equations, but most other historians retain the older tradition of descriptive or argumentative narrative combined with citations. Certainly those who want to write books that lots of people will read.
I’d guess the ideal amount of evidence depends largely on the intended audience (scholarly or popular), which boils down to the question of how interesting the average reader finds the topic under discussion. The assumption is that popular audiences hate footnotes, but we should admit that busy academic historians skim and dip through many academic tomes – seeking out those passages relevant to our particular interests, while leaving the author’s larger debate over whether the aristocracy rose or fell to others more interested in that discussion. Only works that are more directly relevant to our particular combination of period-place-topic merit sustained attention. One result of this is that less popular fields (like EMEMH) will likely feel more pressure to get to the point with the minimum number of details, and/or else expand their scope to more interesting periods.
Perhaps ironically, these pressures to stretch an idea into a more marketable product occur at exactly the same time as digital collections have allowed early modern historians to pull up a chair to the information buffet that modern historians have been gorging themselves at for some time now. Not only has the availability of digitized book and newspaper collections increased the amount of information at our fingertips, it has also increased (rightly) historians’ expectations regarding evidence. Anyone who went through American grad school in early modern European history during the 1990s knows how poorly sourced most of our class papers were, how starved we were for primary sources if we were in a non-American (or non-Ancient) field. Heck, the main reason why I even got into the subject of the War of the Spanish Succession in my first year of grad school was because there were numerous published volumes of correspondence available in the library. And I can’t even count the number of pangs of longing I felt stumbling across yet another footnote citing a military treatise locked away in some rare book collection hundreds of miles away. Oh, cruel fate! Just as we gain access to these works, we’re told we can’t cite them in our books!
So on the one hand, those long forced to live on meagre rations see footnotes as a way to show off the fruits of our obsessive hoarding behavior, footnote as curio cabinet. But beyond this catharsis and exhibitionism, there are more defensible reasons why we, history-reading consumers as well as history content creators, really do need more than one example. How many examples will depend on the type of historical research one is engaged in – some focused research questions (such as a biography or a narrative of a specific event) may only require a single, telling example, whereas questions about mindsets and mentalities may require a broader range and greater volume of evidence. Generally though, I’d argue we need historians to fully cite the evidence in the following cases:
- When a point is particularly controversial in the historiography. Unfortunately, explaining to the reader why a point merits such attention to detail would itself require a discussion of the current historiography, which is one of the elements of scholarship that editors tell us we need to jettison.
- When a fact or view is oft-repeated, but wrong. A surprising number of “facts” in history, often the most well-known, turn out to have pathetically little supporting evidence, and are often flat-out wrong. Yet they get repeated again and again because they are useful, usually because they neatly encapsulate our idea about the subject as a whole. There are probably a dozen in my small bailiwick of early modern siege warfare: that Vauban could accurately predict the length of a siege, that Vauban’s siege attack was easy and slavishly applied…
- When a claim is particularly counter-intuitive. Figuring out what is counter-intuitive gets a little challenging if you’re writing for a broader audience that might not know the ins and outs of your particular subfield, that is to say what’s “obvious” to the experts.
- When we want to know not just whether something was possible, but whether it was a) deemed probable and b) actually more likely to occur. A single quote will tell you whether somebody could imagine something happening, or how an individual interpreted an event. Most historians, however, would really like to know whether that idea/interpretation was plausible. All else being equal, I’d contend that the more frequently expressed ideas were also likely to be those contemporaries deemed more probable, or if you prefer, more contemporaries deemed it possible. [Things become more complicated when we’re dealing with potentially widespread views that were rarely uttered exactly because they were so fundamental. But if you have two competing hypotheses, one with a single supporting piece of evidence and the other with multiple cases, I know which way I’d lean, and listing the evidence for each would sure help.]
Whether such probabilities were actually manifested in the real world is of course yet another step, requiring yet more evidence, usually from a different type of source. In my research field, for example, you need to compare theoretical, prescriptive treatises of how to wage war against narratives of campaigns, and against the less-filtered correspondence written on the spot. Cite your sources so the reader knows you know the difference between potential and actual.
- When we want to know how widespread an idea/phenomenon was. This lets us know whether some historical actor being cited is a genius ahead-of-his-time or someone representing a broad consensus, or whether an event is unique or commonplace. Maybe our argument is focusing on the crackpot genius, but just as often busy historians use a quote to represent a presumed consensus to stand in for an entire period or country or group. Unfortunately we also know that historians, especially when surveying multiple periods/places, rely on easily-accessible, easily-digestable published works, which explains why oft-reprinted military treatises and survey works like The Art of Warfare in the Age of [blank] get cited so frequently. Unless we know how such prescriptive works were received, however, we can’t really tell how widely their opinions were shared. Maybe we can even glean the relative degree of certainty of the different parts of a historian’s argument. For example, does a pivotal point of an author’s argument rely on a single example, whereas subordinate claims have much more evidence? Depending on the type of argument, perhaps the author added an unwarranted co-premise and thereby stretched their conclusion beyond what the available evidence suggests.
- When we want to know whether there’s a pattern underneath. Careful historians won’t assume that one quote summarizes an entire generation’s mindset – people were just as argumentative 400 years ago as they are now. Perhaps an opinion truly was monolithic, but more likely it was only dominant or merely held by a simple majority, or evenly split among competing camps. Or maybe it was a pedantic question only of interest to a few contemporaries. Maybe opinions on the matter varied by political orientation, or social status, or nationality… Maybe what the majority of the English electorate believed in 1704 changed radically by 1711? Maybe the English believed one thing but the French believed another? Multiple examples reassure us that the author has actually looked for points of contention instead of painting with a ridiculously-broad brush. Plus, their citations allow us to see the types of sources they’ve consulted.
- When we appeal to zeitgeist for explanation. This risks committing a fallacy comparable to the appeal-to-authority – “the zeitgeist shaped sphere X because in some other sphere – usually an intellectual/cultural one – they believed in the general principle of Y.” Most obvious in EMEMH would be the question of how far scientific ideals of the Enlightenment penetrated military thought, military organizations, and, more contentiously, military practice. (J.R. Hale, among others, has asked similar questions with regards to the Renaissance – was there a Military Renaissance?) Many military historians are particularly uncomfortable with zeitgeist determinism, since warfare is generally seen as being shaped by the dictates of practical factors more than preconceived notions of how war should be fought – plans not surviving first contact with the enemy and all that. An open question to be sure, and one which requires evidence not of some posited cultural straitjacket, but of specific examples from the “sharp end” of combat as well.
- When we want to restore our faith in the historical profession. Every once in a while I’ll come across a contemporary source saying something that encapsulates a logic that I’ve seen ascribed to contemporaries elsewhere, yet never actually seen another historian quote or cite some historical figure spelling out the logic – this is particularly important given the temptation to fall back on the universalistic “Inherent Military Probability.” Perhaps it’s a bit pedantic of me, but I really want to see contemporaries saying it on the printed page, preferably as a quote rather than a citation. Quoting contemporaries might even help other historians identify important vocabulary (“vigor”, “politick” and the like). Did contemporaries explicitly refer to that battle as “Pyrrhic”, or is that a metaphor utilized solely by the historian? It matters a lot to the historian interested in the dissemination of Ancient ideas.
And I know I’m not the only historian to want to see the evidence. I recall, at a graduate student conference way back when, a young presenter contending that the Irish (16C? 17C?) buried their cannon on campaign because they didn’t want to hinder their mobility. A more senior graduate student in the audience challenged the speaker, asking how she knew that was the reason for the burial and not some other? Her response: “Because they said so.” Audience laughter at the interrogator being put in his place, but his muffled response was spot on: “Well, you need to tell us that.” Historians are from Missouri, witness the frequency with which reviewers bemoan books without this fundamental scholarly apparatus.
- When we want to really understand our subjects. To distinguish the ways contemporaries were similar to and different from us today, we need to immerse ourselves in the period, by reading as much as we can. We want to know that the authors we’re reading have done this, and authors indicate this immersion through footnotes in particular. This is particularly important for researchers who are writing about periods and places clearly outside of their expertise, as often happens when someone explores a topic across several centuries. And just because you’re an established scholar doesn’t mean the rules of evidence don’t apply to you. Feel free to be as impressionistic in your work as you want, but realize that ideas are a dime a dozen – the hard, and convincing, work is to build the argument to support your claims. Contemporary evidence does a lot of the convincing, because historians are trained to always wonder how foreign our historical subjects’ thoughts were from our own.
All these reasons – most methodological but a few professional – shout out for footnotes and citations. But we shouldn’t blindly equate a bevy of citations with an unassailable argument; we need intelligent use of citations, not just piling up a mountain of cites for the sake of it. The digital cornucopia, a veritable orgy of sources, has perhaps paradoxically had the methodological effect of actually making it easy for just about any theory to find at least some support. As they say, you can usually find evidence to support just about any claim if you look hard enough, and the likelihood of finding one supporting example increases as the amount of information grows, much like the R-squared value of a multiple regression model will necessarily increase as you add more variables. So we need to be careful. Have you developed multiple ways to test your theory? Do you have a good sample of different types of sources, from different perspectives? Do you understand how those sources were created, and how their genesis might shape the content? Have you looked for evidence that could disconfirm your claim (or distinguish it from competing hypotheses), as much as you seek out proof that supports it? Have you modulated your claims to mirror the volume of evidence that supports it? Do you balance your research between finding the norm vs. finding the outlier? Have you paid close attention to the language used by contemporaries, keeping a special eye on terms whose connotations might have changed over time, and making sure not to conflate contemporary categories into modern ones? These are basic questions of historical research to be sure, but as our data mountain grows, we need to ask these questions even more insistently.
What about secondary sources?
Notice that I haven’t focused on the question of how we should be citing secondary sources. To my mind, this is more difficult since there are reasons that both encourage and discourage the use of citing numerous secondary sources. Obviously you need to indicate where you are getting your ideas from, although this verges dangerously close to historiography. The broader the scope of a work (geographically and chronologically), the more likely you’ll need to cite secondary summaries instead of discrete primary sources. Unfortunately this too often allows people to only skim the secondary lit and avoid getting to know the contemporaries under discussion. In my own work, the question I most often face is how many secondary works should I cite on a given topic – the (unhelpful) answer is “Enough to show that you know the literature but not so much as to bore the reader.”
On the other hand there should probably be fewer “social” citations, where you cite your friends largely because they’re your friends, or because you want to show off the fact that you know about some philosopher from sixty years ago. There’s also the borderline case of whether to cite a secondary source that itself cites a primary source that you use. My criteria is that if I rely on that secondary author’s version (or translation), or if it’s a primary source that I wouldn’t have thought to look at, then they deserve the credit. But if I go and find the original – especially if I already have a copy of it! – I’ll cite the original after I’ve looked through it and processed it for myself. The AHA backs me up on this; every scholar relies upon those giant shoulders that Newton was talking about.
Personally, I want to see the evidence – if I’m taking the time to look at the book, do me the courtesy of showing me the money. I want to see how contemporaries described the concepts and events, how frequently they discussed the matter, how many perspectives contemporaries held on the subject. I’d even like to know how often they expressed their opinion on a subject compared to the number of opportunities they had to discuss the matter. Impossible for some types of history I know, but a guy can dream, can’t he?
So what does this all mean? As an author you could refuse to compromise your evidentiary principles, which probably means publishing with a press like Brill, that will allow you to go into great detail, with a correspondingly inflated list price. You could keep the most sexy example in the main text and relegate the rest to your notes: “Footnote 12: Twelve other contemporaries from both political parties expressed the same opinion.” Doable, but the formal conventions of citations would quickly turn such notes into paragraphs, which is probably not acceptable to many presses. You could include a barebones scholarly apparatus, and cite all sorts of complimentary examples on your blog or website. Maybe we could rely on data-dense visualizations like other fields do, assuming we can convert the quotes from text to quanta. Other options or thoughts?