Fireside this week! A little break after wrapping up our look at Europa Universalis IV. The next things coming up on the blog are going to be a look at who the Romans were and who they thought they were and an extension of our examination of EU4 into one of Paradox’s other titles, Victoria II. The series on identity in the Roman world is likely to be the one that appears first. After those, I am hoping to be ready to dive into the long-awaited discussion of doctrine and perhaps also trench warfare.
For this week’s musing, I thought I might talk a bit about the relationship between historians and political scientists. I sparked a bit of a teapot tempest on twitter when I noted that at a panel at the annual meeting of the Society for Military History on the legacy of Samuel Huntington’s The Soldier and the State (1957), which still exerts quite an influence on thinking within the military, that the Q&A had turned into a mild and friendly but amusing critique of political science as a discipline:
To be clear, the general consensus of the discussion tended towards, “useful interlocutors with whom we will frequently disagree.” I’ll admit I was more than a bit surprised that some political scientists took umbrage; it has been my experience that any meeting between political scientists and historians involves at least some gentle ribbing by both parties about the differences between the disciplines (more so if alcohol is served at the event). But then it occurred to me that the relationship between the two disciplines is exactly the sort of inside baseball that these firesides are for.
So let’s talk about it: why is there a cordial tension between political science and history?
History and political science were bound to be interlinked disciplines, because they generally work with some of the same evidence base (historical case studies derived from the evidence of the past) and are often interested in the same questions (how polities function); where they differ most is in methods. Fundamentally, history – the older of the two disciplines by approximately 2,300 years – is a discipline of the humanities and applies humanistic methods (like close-reading, textual analysis, and so on), while political science attempts to answer questions about how polities function through scientific – or if you are being less generous, ‘scientific’ – methods. It would be a bit too simplistic to say that history is a ‘words’ discipline and poli-sci is a ‘numbers’ discipline (some historians use statistical methods, many political scientists focus on theory which is often not numerically quantified), though overall there is more emphasis on arguments which can be quantified or expressed in equations in political science than in history.
That methodology though, is in service of different goals. Political science is fundamentally oriented towards a search for generalizable conclusions – things which are true not merely in a case study, but about many societies more generally and may thus be applied out broadly. History, by contrast, as a discipline tends to focus on contingency, which in turn minimizes (without fully rejecting) the possibility of generalizable conclusions. To be a historian is to accept that, for instance, Greece and Rome are quite different and that conclusions from one society often cannot be safely generalized to the other (much less to say, early modern France). Moreover, even in societies that are very similar accidents of chance and local conditions can still mean that conclusions in one society might not generalize to another. Because of that focus, historians tend to operate at a much higher level of granularity – we tend to ‘zoom in’ a lot more – which in turn means that our studies tend to be consumed by this non-generalizable contingent elements of history.
All of which produces a very predictable cycle. First, historians examine the primary source evidence for past societies and produce processed historical interpretations of them (that is, they produce history from evidence). Though the number is by no means zero, relatively few political scientists interact with the raw evidence of societies even modestly deep in the past (though this is not the only form of evidence they use, see note below); all historians do, by definition. Consequently, the built up history – which is again, a processed good produced by historians out of evidence – is then used as a foundation for political scientists to produce research arguing for a generalized conclusion about societies (often embodied in a mathematical system which by necessity has taken irreducibly complex societies and reduced them to shorthand formulas). And then, with all the constancy of the North Star, historians cry foul at the generalized conclusion, producing from their very focused research on the idiosyncrasies of various societies a thousand and one exceptions to the generalized principle (which may or may not, in the event, actually overwhelm the value of the principle; even good rules of thumb have exceptions, after all).
(A word on political science’s evidence base for a moment. Obviously, historian-intermediated history (or historian-intermediated ‘data,’ but I think we should be cautious with that word – ‘data’ implies a clear truth value which the historian in me is reluctant to offer) is not the sole source of the ‘data’ that underlies political science work. A significant minority of political science is experimental – attempting to replicate the actual scientific method – though this remains very much a minority because controlled experiments in things like war or tax policy are difficult or impossible. A lot more political science relies on using modern administrative data produced by modern states. The tricky thing there is that this data doesn’t go very far into the past, relatively speaking, before it needs a historian to mediate it or risk catastrophic mistakes of interpretation. Consequently, for societies much older than the present (really, anything much before 1900), political scientists are often substantially reliant on historians to create their data from the historical evidence. Finally theory and behavior studies from the sciences proper also often serve as evidentiary foundation for political science studies as well, much as they do in some of the other social sciences)
This relationship is compounded by meaningful differences in the standards of evidence and argumentation between the two disciplines. Historians are generally trained to reject simplification and abstraction; you aren’t supposed to paper over exceptions, you are supposed to dutifully note every last one in your footnotes. By contrast, the goal of political science is to extrapolate general rules which by necessity must simplify substantial variations in the evidence in order to produce rules of thumb on which policy may be based. Likewise, historical argumentation comes with a greater emphasis on historiography – the study of the history of the history. There is thus a general expectation that a historian of a topic is familiar not only with the topic itself, but with the decades or even centuries of historical writing on that topic (the history of the history of a topic, as it were). Political science, by contrast, isn’t about documenting the past but detecting systems for the present and so tends to be quicker to ‘deprecate’ (in the software sense of ‘declare obsolete’) older scholarship; this isn’t a universal tendency, some political scientists are very good at assembling the ‘history of the debate’ but the pressure to do so tends to be somewhat less (but with a greater demanded awareness of contemporary theories). Consequently, when historians cry foul, they often do so in evidentiary terms which aren’t as applicable with political science, leading to frustrations all around.
On the flip side, by political science standards, a lot of historical studies go absolutely nowhere. A micro-history that studies a small town in the 19th century and concludes that its social organization is entirely sui generis and cannot be meaningfully extrapolated to anywhere else is a potentially really interesting historical study, but not a particularly exciting political science study. Because history is in part focused on documentation (paired with interpretation), we’re a lot more willing to spend a bunch of time documenting things which, in the event, were only significant to a handful of people. I have, on this basis, heard political scientists describe this side of the historical method as ‘boring.’ Moreover, historians are often quite resistant to reducing their conclusions to data, especially numerical data (we are, after all, scholars of the humanities which tends to make us fundamentally word people).
And to be clear, this is fine. Different disciplines have different methodologies. Pure ‘death of the author’ literary analysis would also be a hard sell in a history context (because providing the context for the production of a work is what we do) but is perfectly appropriate in a literature setting (or historical study focused on modern reception). But because historians and political scientists tend to play in the same pond (particularly with political and military historians), those different methodologies tend to interact rather more, which leads to some fairly predictable friction.
Now clearly I am not impartial between these disciplines; I have picked a side. When it comes to the study of historical polities, I think the humanistic method is more likely to yield useful insights than a political-science method (especially in the ancient world). Indeed, I often think that the aspiration to be a science frequently holds political science back; I have already written about how I think some problems are not susceptible to the scientific method. And especially when it comes to the ancient world – where the actual evidence is often far weaker than a quick browse of the secondary scholarship would imply – I often find myself in the historian’s typical position: the ‘you can’t know that’ guy pulling the brakes on Big Arguments fueled by poli-sci data-centric approaches. Or alternately, the annoying fellow whose ancient evidence base contradicts pleasant theories constructed on databases that don’t extend back before 1500 (e.g. the ‘democratic peace theory‘ which works more-or-less with a modern dataset but collapses when ancient examples are included; representative ancient governments went to war with each other all the time). Generally, I’d say the modernist bias in political science is a real weakness; when pre-modern evidence is used it is often used with startling lack of nuance and care for the complications it typically poses.
The conversion of ‘evidence’ into ‘data’ is also often a violent and distorting process. Databases are only as good as what is included in them, as neatly exposed by a recent paper pointing out that Native American conflicts had been almost entirely left out of most major Poli-sci conflict databases; but then that study has its own flaws because its neatly graphed figures are heavily distorted by the evidentiary bias to report conflicts between Native American and colonial powers and then not necessarily report warfare that did not involve colonial powers (in some cases because such conflicts leave limited evidence).
(Also, on the educational side, I have also found that some – by no means all! – political science programs teach their students to write in the rigid pattern and language of a scientific paper, a form sometimes adopted by political scientists (often where appropriate, sometime less so). Of course that style of writing has its purpose (especially in the STEM fields), but teaching that style in lieu of traditional persuasive essay writing is doing many students a disservice. Even a political scientist isn’t likely to use that rigid format outside of narrowly academic environments and most students in those programs aren’t going to go on to do the sort of poli-sci research which demands that kind of writing anyway. I do not know how general a phenomenon that is in teaching, but I seem to encounter students trained this way with some regularity (and never from the STEM fields narrowly construed).)
All of that said, political science produces useful insights. We have, in fact, just been discussing the neo-realist theory of Kenneth Waltz, a political scientist, which I find a very useful frame – not the only one mind you, but a useful one – for understanding the behavior of states and interstate systems. Likewise, Stephen Biddle’s classification of the ‘modern system of war’ is a work of political science and while I have some very strong quibbles with how he attempts to assess and express that theory mathematically, the overall argument is sound. Moreover, the interaction whereby political scientists suggest grand conclusions and then historians cry foul is also a valuable knowledge-generating process, assuming that one is reading both the original grand conclusion and the objections (too often, it seems to me, writers aiming at a popular audience read only the original, grand conclusions and remain blissfully unaware of their reception by historians). Ideally, the best of those grand conclusions survive the process and become foundational theory which both political scientists and historians can draw on when trying to understand events and societies in the past.
In any event, this relationship goes a long way to explaining the amiable competitive tension that often exists between the two fields. I was rather surprised by a number of political scientists on Twitter who seemed caught entirely unawares by the existence of that tension, since I have found that it emerges quite quickly any time historians and political scientists get together. I rather wonder if some of the folks caught off-guard by that might do better to hang out with more historians.
On to recommendations!
Drachinifel has a podcast-style interview with Jon Parshall (co-author on Shattered Sword which I have recommended in this very space) talking about the Battle of Midway and the American naval position in 1942 in the Pacific more generally. Parshall is an expert on Midway and 1942 in the Pacific and Drachinifel – whose channel is all about warships in the late 19th and early 20th centuries – asks good questions, so the resulting discussion is really quite good, especially where they get outside of the four corners of Shattered Sword and start discussing the broader context of the events of the year.
While we are on podcasts, I should note that I was on a podcast with Murray, Jasper and Mark over at Ancient Warfare Magazine‘s Ancient Warfare Podcast. We discussed one of the topics of my current research, the Roman adoption of mail armor in the third century and its impact in the second century BC. It was a good, fairly freewheeling discussion (we occasionally got a touch off topic) and is well worth a listen. The lorica hamata, introduced in the third century to Rome (invented by the Gauls in the Danube River basin) is a lot less talked about and focused on compared to the more iconic segmented plate ‘lorica segementata’ (it is worth noting that while lorica hamata, ‘cuirass of hooks’ is actual classical Latin, lorica segmentata, ‘segmented cuirass’ is never used in our sources and is instead a modern invention terminology wise). But the hamata seems to have been substantially more common, in use by the Romans from the late third century BC through to the end of the empire in the West some seven centuries and change later. Unlike the segmentata, is also survives as an extremely common armor-form throughout the Middle Ages.
Meanwhile, Season 2 of Peopling the Past‘s podcast series is starting up, focusing on Roman Art and Archaeology. They begin with a fantastic discussion of some of the truly incredible remains at the Roman fort of Vindolanda with Dr. Elizabeth M. Greene, where the fairly unique local soil has preserved all sorts of things which would normally never survive in most archaeological sites – in this case, literally hundreds of shoes. And again, I make this plea: if Peopling the Past is not on your normal blog list, do go and check it out – each of their posts features classical scholars doing really interesting projects. For instance, in late April they featured Dr. Christine Johnston talking about Egyptian lifeways and the influence of the Nile on them, complete with some processes of production which ought to seem mighty familiar to readers of ACOUP.
Finally, book recommendations! This week, I am going to recommend I. Haynes, Blood of the Provinces: The Roman Auxilia and the Making of Provincial Society from Augustus to the Severans (2013). While the Roman legions get the lion’s share of the press for Rome’s armies, beginning in the Late Republic, Rome began recruiting units of allied non-Italians to fight in units that became known as auxilia which were standardized under the reign of the first emperor, Augustus (r. 31BC – 14AD) and by his death made up approximately half of the Roman army. These auxiliaries were paid professionals recruited from the provinces (generally from non-citizen provincials) organized into cohorts (of infantry) and alae (of cavalry) which might be deployed alone or attached to the (larger) legions, who were rewarded with grants of citizenship on retirement. Often, as Haynes notes, more flexible tactically than the legions themselves, they both screened the legions in combat and sometimes carried out the battle entirely without the intervention of the legions.
Haynes goes behind just the battlefield role of the auxilia though, laying out the evidence for their recruitment, the conditions of their military service, the quirks of religious practice, their equipment and so on. This aspect of the book, that it is an effort at a really complete study (in contrast to older books on the topic like D.B. Saddington’s The Development of the Roman Auxiliary Forces from Caesar to Vespasian (1982)) means that the book touches on a lot more than tactics and combat (though, again, it does address equipment, tactics and combat!). The date range is also welcome – Haynes efforts, in as much as the evidence allows, to follow the institutions of the auxilia from their formalization under Augustus all the way to the Severans, where the grant of citizenship to all free person in the Empire (the Constitutio Antoniniana (212AD) – try having to say that at speed during a lecture) made the institution fundamentally redundant. That advantage of date range comes out as Haynes is also to identify changes, such as the tendency for the ‘ethnic’ quality of auxiliary units, in equipment, fighting style and identity, to wane over time.
The book is not short – 382 pages not counting end-matter (and priced to match) – but while it is not really written for a popular audience, it is written quite well and easy to follow. The length allows Haynes the virtue of taking a good deal of the evidence, both from the surviving literature of the period but also a lot of epigraphic and representational evidence (the book is furnished with many good images, especially of funerary artwork showing auxiliary soldiers), and putting it directly in front of the reader, so you not only have his conclusions but you understand the kind of evidence that exists as the basis for them. And while the book is by no means dumbed down, it is also quite easy to follow for a non-specialist; key terms are translated and I think a regular reader will require only minimal wiki-walks to keep up.
Frankly, for anyone looking to understand imperial armies which make use of provincial manpower to fill their ranks – both how they do that but also what that is like for the folks in question and how those soldiers relate to the structures of imperial power they are joining, reading Blood of the Provinces and last fireside‘s Soldiers of Empire together will put your feet firmly in the real (a handy thing both for historical interest, but also for worldbuilders looking to make empires that actually work!)