“That bad, huh?” the brother said when I slapped down Phil Sabin’s Lost Battles on the table in the campus Starbucks as he joined me for coffee yesterday. Actually, it is not bad at all – far from it. It is a good book, the product of many years of careful scholarship, and a useful contribution to both military history and game design and cliometrics. Unfortunately, I know enough about these to have opinions which differ from Sabin’s and which make me want to argue with the book and I’m going be annoyed that I will never have enough time to run down enough sufficient evidence to test my ideas properly against his. Obviously, this is not going to be a proper review, but more of a first impression. So far (and I accept that I may be proved embarrassingly wrong later in the book) I think he is wrong to ignore articulation and wrong (in the polite academic sense, not the football namecalling sense) about hordes of cavalry.
What Sabin has done is to offer a scalable model to interpret ancient battles at the grand tactical level. That the model is scalable is important – it works for battles from small fights between Greek city states right up to the monsters of ancient warfare like Cannae. Since it is scalable, it allows for useful generalizations about the nature of ancient warfare. However, in order to make it work, he has had to impose a level of abstraction on army organisation in the battles, standardizing each army at about 20 units with differing fighting worth. Now, this is essential if you are going to compare different historical events with vastly different real world magnitudes in a way that allows you to make meaningful statistical comparisons. The problem is that once you chose a modelling methodology and impose it on historical reality, you chose to abstract out some variables, and this method puts military organisation to one side. This, I think, is a shortcoming which could lead to problems since one of the things which distinguishes armies right across history is articulation – the extent to which they were organised into units which allowed the commanders to effectively manoeuvre them and literally articulate his command intent on the field. Right through the history of warfare, some unit sizes appear again and again because they work – Macedonian taxis, Roman Cohort, Swedish squadron, modern battalion. The 600 or so strong unit is a clear building block of successful armies throughut history. Not all successful armies have had it, but almost all armies that have, when matched against armies without effective articulation, tend to win. It is not something I like to see abstracted out of an combat simulation, and of course Sabin does a lot to factor it back in implicitly under the heading of command in Chapter 5.
On cavalry, he argues that “Polybius’s generalzation about the importance of cavalry superiority is in fact starkly at odds with the outcomes at Platea, the Granicus, Gaugamela, Magnesia and Pharsalus, where in every case the side with far more horsemen suffered a resounding defeat” (p 22) Polybius is indeed often wrong, but what Sabin seems to ignore is that in many battles where one side had vast numbers of cavalry, that was light cavalry, which was often useless against a properly handled mass of heavy infantry. Indeed, the role of heavy infantry as the ‘rock’ in Archer Jones’ 4 way version of rock, paper scissors is a vital part of that model. That is a a model which Sabin discusses on the next page, and ends by telling us that he is going to categorise troops as heavy and light but intends to ignore the difference “..because of the ‘swings and roundabouts’ nature of the relationships involved.” Somehow I feel that when I get deeper into Sabin’s model of ancient warfare, I’ll find that Jones’ four-way model of heavy and light, infantry and cavalry will force its way into the calculations.
So you see it is a book which is going to provide a great deal of argument value over the coming years. Creating a model which allows us to re-interpret a range of battles from Marathon in 490 BCE to Pharsalus in 48 BCE is a serious task. I’m sure that when I get into the model and have a go at fitting it to other ancient battles, I’m going to have a bit of fun, and that is one of the main reasons why I teach history. Even though my military history teaching is all officially post-1450, I am pretty sure I can fit some of it on the reading list and hopefully the undergrads won’t be too lost when I explain why my Argyraspids are actually better than Phil says they are!
Sabin, Phillip Lost Battles: Reconstructing the Great Clashes of the Ancient World, London, Hambledon Continium, 2007 ISBN 9781 84725 187 9
Leave a Reply