Fireside this week! We’ve just moved and I am settling in to my new home office, but the slow process of unpacking all of my books has delayed Logistics, Part III. I can, however, give you a picture of the new Fireside, albeit unlit because it is Augustus in the Carolinas and that means it is hot. Nevertheless, blog readers one and all, BEAR WITNESS:
For my musing I am going to be a bit silly and talk about a common science fiction setting, the post-apocalypse, and what those conditions might mean for warfare, at least in a very simplified way. In part this is because a lot of the outcomes that appear in the genre – particularly in games with a focus on combat – don’t seem to me to be likely or stable outcomes, regardless of the sort of apocalypse that has occurred. We’re going to assume a Fallout, Dying Light or Metro style apocalypse; the specifics are unimportant but the idea here is that enough time has passed for new societies adapted to the new conditions to emerge and there are enough humans to make more or less stable societies possible. And in part this is because I actually want to talk about what it takes to make a modern army able to fight because even the things that seem very simple are in fact very complex when you cannot just order them from a convenient supplier.
In the immediate aftermath of whatever disaster has happened, these societies – with vastly fewer humans than they started with – are going to initially have a lot of firearms (there are something like a billion guns worldwide) with a much smaller population and a lot of modern ammunition that is already accessible. But in conditions where there is a lot of conflict, that ammunition isn’t actually going to last all that long. Modern infantry tactics rely heavily on suppression fire to enable maneuver, which means expending a lot of bullets – even for armies that prize marksmanship and accuracy – in any given fight. This is something that doesn’t seem to have filtered through to the general public very well; in the more intense days of the Global War on Terror there were occasional stories about how many hundreds of thousands of bullets were fired for every enemy combatant killed, printed without much of an awareness that this was the intended use case. Bullets are cheap, soldiers are expensive, you fire a bunch of the former to try and avoid losing any of the latter.
That style of fighting isn’t a fashion choice, rather it is the most effective way to stay alive and defeat an opponent in a firefight when you have a lot of bullets. But it is going to rapidly result in having far less bullets or more correctly far less cartridges. And here we need to get into the different parts of the cartridge, the thing a modern firearm fires. Generally speaking there is a projectile (the bullet), a propellant charge, a primer (which ignites the propellant charge when struck) and a casing to hold it all together. The cartridge (technically here the ‘integrated cartridge’) was actually a major innovation in warfare; in early firearms all of these elements were separate and had to be loaded in order to fire. Even for a single-action firearm, being able to just load the cartridge and get bullet, powder and primer all in one was a substantial improvement.
The good news here is that the bullet is a fairly simple lump of metal and casings are both reusable and not too difficult to machine or cast, so a small community of survivors (think the size of a small town) or even a particularly handy individual could easily produce their own bullets and casings. Indeed, some modern reloaders do exactly this. But primer and propellant are going to be a problem.
Modern propellants, sometimes collectively called ‘smokeless powder’ are chemically complex things, a product of the rapid innovations in chemistry in the 1800s. The primary component of many modern propellants (though not all of them, some use nitroglycerin1) is nitrocellulose also called (quite sensibly) ‘guncotton’ or ‘flash paper.’ It is tricky to make; the first effort to manufacture it at scale in 1846 blew up the plant within a year and it took 15 years from that to figure out how to manufacture it safely without it just exploding. Seriously, looking into the history of guncotton is a succession of plant explosions (Faversham in 1847, in Austria in 1862 (and 1865!), and Stowmarket in 1871). To make it, you need nitric acid (HNO3) and sulfuric acid (H2SO4) and a source of cellulose (C6H10O5 over and over again); any form of cellulose will do (but the quality matters); the first guncotton was made using cotton (imagine that). Drying must be done very carefully and also the ratio of nitric acid to sulfuric acid has to be controlled carefully; too much nitric acid and the dry guncotton explodes near room temperatures, which is not ideal.
But that’s not all you have to do here, because additional processing is going to be necessary to get the propellant into the right form (grains or filaments) and stable enough for use in a rifle round. This required yet more additives and produced various formulas like ballistite and cordite (both of which use nitroglycerin in addition to nitrocellulose), with the nitrocellulose in some cases coming in the form of collodion (a solution of nitrocellulose, ether and alcohol which requires its own chemical process to produce). And on the other end of the spectrum, all of these 19th century chemists could rely on just being able to order their precursor chemicals, but actually producing nitric and sulfuric acids (both required for nitroglycerin and nitrocellulose) is also tricky and dangerous. These are, after all, very powerful acids.
In looking this all up I fear I may have gotten my name on some sort of a list. Also I cannot tell you the number of times I heard or read some version of ‘do not, under any circumstances, try this at home without <safety device you absolutely do not have at home> and <training you absolutely do not have>.’ Given just how many inventors appear to have blown up themselves, their factories or close relatives trying to figure this stuff out, I am content merely to read about it. But, in case I am not being clear: do not try this at home. Or anywhere.
The point here is that making a modern smokeless powder at even a small scale – even older, less chemically complex kinds – is going to require at minimum a decently well-furnished chemistry lab. Producing this stuff at any kind of scale is going to require a chemical plant and remember, you want hundreds of rounds to be able to fire in order to do modern system infantry tactics; a standard US soldier’s ammunition load (210 rounds of 5.56 NATO) requires about 6,000 grains of propellant, 0.85lbs or so. Per combatant. In either case, you are also going to need a chemist who understands the processes and is either experienced in or can think through the necessary steps to stabilize the resultant highly unstable explosives without them exploding. It isn’t impossible to do this, but it is going to be a struggle for a post-apocalyptic community, demanding both knowledge and the ability to support specialization; doing it at scale also demands a lot of infrastructure and remember that you need to do it at scale if you want to use the more effective modern tactics which require putting a lot of metal down range quickly.
And this is actually a neat point to come out of this little exercise: the tremendous complexity of producing even the simplest elements of modern industrial military power. If you want to build a modern army, you must start with a chemical plant (actually quite a few) or else purchase your ammunition from abroad (and thus be strategically captive to your suppliers because, again, in an actual fight your ammunition stores will not last. This ‘access problem’ is actually older than smokeless powder; we see European powers exploit it against local powers in both West Africa and North America).
Does this mean that guns drop away until eventually it is back mostly to bows, arrows and contact weapons, like in Dying Light 2 or Horizon: Zero Dawn? No. Because there is a simple, backyard producible alternative to modern propellants: black powder. Black powder is made of saltpeter (potassium nitrate), charcoal and sulfur in a roughly 75/15/10% mixture; getting the mix a little wrong will produce an inferior propellant, but any mixture with more than 60% potassium nitrate will explode when ignited. Charcoal is just a mass of carbon, easily producible in sufficient quantity and sulfur is a fairly readily identifiable naturally occurring mineral and one of the most abundant on Earth. And saltpeter, while it occurs in mineral form can also be produced from excrement using a nitre bed to encourage nitrification, a chemical process one can literally perform in their backyard using a product (human waste) that human settlements produce automatically. And figuring this out isn’t necessarily going to require an expert chemist either; your average town library likely has multiple books which describe the production of black powder from a historical perspective.
Now black powder is, as a propellant, markedly inferior to modern smokeless propellants. It produces lower pressures resulting in lower muzzle velocities and thus is going to produce inferior range and penetrative power for a given size of cartridge and gun. It also doesn’t burn very cleanly; that’s what makes it not ‘smokeless’ compared to modern powders (which also still produce some smoke, just a lot less). In modern, rapid-firing firearms, the half-burnt material black powder is going to leave behind will quickly foul the weapon. But, assuming you can ignite it, it will fire and it isn’t hard to see survivors in this situation quickly identifying which sorts of firearms had actions which could at least tolerate the inferior powder they could actually produce in bulk rather than requiring the limited and dwindling supply of modern cartridges or bespoke chemist produced propellant.
That leaves primer, which is also going to be tricky, but on the upside you need a lot less of it. The key here is a chemical which will ignite when struck by the firing pin (and so ignite the rest of the charge). The earliest chemical used for the purpose was mercury fulminate, the production of which requires mercury, nitric acid and ethanol. Another early option is potassium chlorate, produced from sodium chlorate and potassium chloride. Once again as best I know we are again looking at products that are going to require a chemistry lab, especially because you need to remember that a community of post-apocalypse survivors can’t simply order the precursor chemicals from a supplier. Still, if your community of survivors has the chemist and the facilities for small-scale production, as far as I can tell batches of primer are going to be more useful than batches of propellant: there is a cheap, backyard (inferior) solution for propellant, but not – as far as I know – for primer. Getting something that is stable under normal conditions but ignites reliably under pressure is a tricky chemical problem, after all. And absent a workable primer, none of the firearms you have will work.
And all of that is going to impose some shaping constraints on warfare in a Fallout style apocalypse. Unlike in many of these settings where guns are scarce, it would be cartridges that would be scarce (credit, I suppose, to Metro, where they are in fact so notionally scarce they’ve become the de facto currency); the world’s supply of guns are more than capable of shooting the world’s supply of bullets long before they all wear out their barrels and actions. These modern weapons are designed to be the last step on a very long industrial pipeline continually feed explosives to consumers (civilian or military); the gun is not itself so much the final product as the let step in the assembly line. At the same time, and I want to stress this, these ammunition-heavy ways of fighting are simply the most effective way of using these weapons, so a community can’t find its way around an ammunition shortage; it simply has to drop back to less effective ways of fighting.
But because black powder isn’t nearly as hard to make (though I must note that many early black powder experiments also ended up with unintended explosions) and modern machining tools can produce basic firearms fairly easily (and there are lots of firearms generally), I doubt you would see a reversion to swords, spears and bows. Instead I suspect you would see tiers develop between various weapons, ammunition and communities. At the upper end, you’d have modern cartridges which would probably end up being hoarded for use with automatic weapons for war and emergencies, since automatic weapons would see the greatest effectiveness loss from messy black powder cartridges (since rapid fire would foul them very quickly). Then you’d have everyday personal defense and hunting weapons, probably using black powder cartridges (with the limited supply of primer being the focus of what chemistry is available to these societies). And if primer became scarce, you might see the reemergence of other jury-rigged ignition systems, potentially in muzzle-loading configurations because again, guns, bullets and black powder would all be fairly easy to produce given a town library and a high school’s shop classroom. After all, you can hunt with a black-powder fed, flint-lock rifled musket, but trying to use that weapon in a firefight is a poor plan unless there are no other options.
The odd difference here is this would be a society that knows what the future looks like in terms of firearms: they know that industrialized warfare is possibly and they know how to wage it if they could only get the means. Consequently – and this is bad news for the plot of almost every post-apocalyptic setting – the advantage is going to go to large coordinated communities which can most rapidly reassemble industrial chemical processes. Probably that would begin with improvised chemical labs before scaling up production into large chemical plants. Communities that were able to scale up production sufficiently to provide for automatic firearms and light artillery would enjoy an enormous advantage over their neighbors (or zombies) almost entirely regardless of numbers. A community able to build the infrastructure for large scale primer and propellant production would find itself with an immensely valuable military resource in a landscape of communities without that capability.
I actually suspect you’d see similar processes play out in several different fields at once, all related to the ability to produce the chemicals required to run modern technology: effective gasoline replacements (which in most cases I suspect would be ethanol) and fertilizers (conveniently overlapping with explosives because the question here is producing nitrates) would all be staggering advantages. We haven’t touched artillery but it presents the same problems here (indeed with many of the same chemicals) only at a much larger scale and with potentially much greater impact. Without access to those replacement chemicals, the wasteland ‘raiders’ and ‘bandits’ of most of these settings would be fairly easy prey to a community which could produce these chemicals. In short the sort of larger town-and-some-farmland community that is normally the prey in these settings to extortion by ‘raiders’ would in fact have the strong hand, able to manage the production of ammunition and thus firepower.
All of which is perhaps a very long-winded way of saying that both House and Caesar’s Legion in Fallout: New Vegas are, in the long run, doomed (just to use a now famous example of this interaction in fiction; New Vegas is absolutely my favorite modern Fallout): the New California Republic (NCR) has very clearly and explicitly hit the point of manufacturing their own industrial firepower – not merely guns and bullets, but artillery as well (and they even maintain and fuel a few pre-war air assets). That means actually destroying NCR is probably already decisively out of reach for the various powers we meet in the Fallout games (short of finding a whole pile of old nukes, a la Lonesome Road) and in the mean time every year the gap between NCR’s growing industrial arsenal and its opponents is going to grow. Also the fact that NCR eventually beat the Brotherhood – a large rebuilding society beating a small, insular ‘tech-hoarder’ society – seems right as well. As we’re seeing, for instance, in Ukraine, at some point advanced firepower is not substitute for raw quantities of firepower. Meanwhile, Caesar’s Legion’ lost before they ever got to Hoover Dam by failing to focus on building up the industrial base to support industrialized warfare.
A final note: the one thing I expect you would see very little of in this environment? Improvised melee weapons made out of random scrap. Fallout and Dying Light are practically flooded with this sort of thing; see also The Walking Dead‘s iconic barbed-wire bat. If you did see a shift back to contact weapons, people are going to want good contact weapons. They’ll likely have access to quite a lot of already acceptable quality steel (and some good spring steel); the fellow with a properly balanced machete or sabre is going to have a huge advantage over someone wielding a repurposed car part or street-sign or whatever. Forging or even just machining a proper weapon would be very doable and even if you are repurposing scrap metal you can always file off extraneous extra bits which might make the weapon heavy or throw off the balance. Beyond the initial phases of the disaster, there is simply going to be no reason to resort to a bat wrapped in barbed wire or with nails through it; you’d be a fool to fight with that. For what it is worth, I’d expect double-purpose weapon/tools like spears, machetes and hatchets to be broadly favored, being robust and useful at a variety of tasks (and thus saving the weight of carrying multiple tools). Spears in particular are potent hunting weapons and their reach offers a huge advantage over most other contact weapons in a lot of situations.
But the upshot of this whole thought experiment is as a useful example of how modern warfare is frequently as much a conflict between industries and economies as it is a conflict between soldiers.
On to this week’s recommendations!
This is becoming a regular recommendation but I want to note that Michael Kofman has once again checked in with Ryan Evans to update the War on the Rocks podcast on the war in Ukraine. I find these updates essential listening as the war has become very slow moving and granular and so I need an expert who can follow what is happening and offer some sense of the state of things. Kofman is also a pretty sober analyst which is valuable in an environment where every new weapon or twitter thread is hailed as game-changing by journalists who are trying their best but may not have a background in military affairs or Russia or Ukraine and so don’t have a sense of what might matter and what might not.
Two links this week from Dr. Eleanor Janega (@goingmedieval on Twitter). First is this video with History Hit on What Was Life Really Like For A Medieval Peasant? A number of the elements will be familiar to those who have read the farming series (small plots, village living, vertical and horizontal ties, big landowners and little peasants, etc) but there’s a lot of additional detail here, some of it specific to the Middle Ages and the whole thing is well illustrated and presented. A lot of history material on YouTube is basically garbage so I’m glad to see History Hit get an actual expert to talk about the topic and the video is really well put together, clear and historically grounded.
Speaking of which, Dr. Janega also let loose on the creators of the new Game of Thrones spin-off series for trying to hide behind the Middle Ages just being ‘rapey’ as an excuse for having lots of sexual violence in their show. The essay is a valuable corrective. I do have one quibble which is I am not so confident that “Early Roman law documents don’t suggest that forced sex was punished at all.” That’s not wrong but the phrasing there is particular: we don’t have a lot of early Roman law (our picture of law in the Late Republic and Empire is much more complete), so the documents don’t suggest mostly because we don’t have the documents. The key foundational law code, the Twelve Tables, survives only in fragments; none of the fragments we have concerns sexual assault but given how little we have that doesn’t mean it wasn’t there. We certainly know that later Romans imagined that sexual assault demanded violent retribution by the injured woman’s family, as evident by legends (possibly based in some fact) about Lucretia and Verginia.
Early law codes often formalized (and constrained) these sorts of customary ‘blood feud’ responses, so it would be very strange for the Twelve Tables not to codify this, though it almost certainly would have been codified the way it was in many other ancient and medieval societies where sexual assault was a crime against the male guardian of the woman in question, rather than the woman herself (which, to be clear, is a rubbish way to set up the law, I’m not making excuses for Roman patriarchy). If it wasn’t so codified, then the attitudes of the Romans as they become historically visible to us (and which they project backwards historically) suggest that fatal retributive violence was considered an acceptable response. As Janega notes, when we finally do get evidence for Roman laws on the matter, sexual assault was a capital offense in Roman law (but as a property crime against the male guardian because Roman patriarchy), which seems a fairly straightforward extrapolation of this attitude. Thus I suspect that the Romans always maintained a fairly severe prohibition against forcible sexual assault.2
Finally this week’s book recommendation was just a very easy one. I’m going to recommend N. Guillerat, J. Scheid, and M. Melocco, Ancient Rome: Infographics (2021). The book is, as the title suggests, a collection of infographics alongside text description covering the whole of the Roman world; the text is fairly chunky and written at about the level of an introductory textbook, though many of the graphics are (because they compress information) actually rather more advanced; for instance the chart of the individual imperial legions with all of their symbols and some service history is the kind of thing you might get in a dedicated Roman warfare course. Text and graphics are divided into three chapters, on one people (including social class, gender, etc.), one on government and religion (including some wonderful graphics of the political structure and processes of the Republic) and one on the military.
Scheid and Melocco are both Roman historians (Scheid, indeed, a senior scholar of no insignificant impact, with a focus on religion; Guillerat’s expertise seems to be on infographics themselves) and their expertise shows here – this is not some half-baked poorly researched book; the graphics are detailed (sometimes stunningly so) and generally very accurate. I did find I had some quibbles with the graphics in the Roman military section – it was here that there were the most curious choices. The fighting formation of the ‘Republican period’ is shown as three lines without the quincunx as opposed to the post-Marian legion; realistically the Republican (449-108) legion needs to be broken at least in two and in practice we ought to admit we have very little firm sense of how the Romans fought pre-225 or so. The ‘number of troops’ graph wants to make a (correct) point about the rising size of the army over time, but it ‘sands off’ the huge mobilization from 218-202 BC to do it – this should be a huge bulge easily visible in the chart where the Roman army jumps to a strength close to what it would have under Augustus (auxiliaries included) before thinning back down for the early second century. The same chart also suggests the Romans were seasonable campaigners in the Middle Republic, which is probably incorrect. Later the quadrireme is suggested as the ‘most common’ Roman oar configuration, which is not what either our sources3 or the Egadi Rams4 tell us (though they currently disagree with each other so you have your pick!).
This sounds like a lot of complaints but actually I think on the whole the quality of the graphics is very good and the information they convey is generally very accurate. There is a lot in this book so I was bound to disagree with some scholarly choices here and there; none of them struck me as a head-into-desk ‘what were they thinking’ sort of the kind I so often see in Roman military books shoveled out to the public riddled with errors. Overall this is a lot better than par for easy-to-absorb visual data, produced by actual historians who actually know what they’re doing running through probably several university courses worth of raw data on the Roman world. The volume itself is hardback, the illustrations are excellent and all at a very reasonable price for the quality of presentation. A great thing for a Roman history enthusiast to have in their library.
- Successfully figuring out how to manufacture that without exploding is what made Alfred Nobel rich and famous, but only after an explosion trying to produce the stuff killed his brother.
- – of Roman citizen women in peacetime, of course. This is a necessary caveat. Roman slavery didn’t acknowledge any sanctity of person until quite late in the imperial period (and even then it seems to have been enforced only very rarely) and as we’ve discussed recently Roman behavior towards ‘enemy civilians’ could be very brutal.
- Which prefer the quinquireme as the most common
- Which seem to suggest that triremes remained common. By the by, if you are at all interested in ancient naval warfare, Stephen DeCasien, the author of that link, is someone you should be following.