Collections: Fortification, Part V: The Age of Industrial Firepower

This is the final part of a five part (I, II, III, IV) series covering some of the basics of fortification, all the way from ancient city walls to modern anti-access/area denial systems. Last week, we looked at the changes which gunpowder-based weaponry enforced on fortification design in Europe, leading to the emergence of the trace italienne.

This week, we’re going to look at the impact first of industrialized firepower (and then briefly at the end, modern precision weaponry) has on the design and purpose of fortifications. In practice, the relative stability of design in trace italienne forts between 1550 and 1750 reflected the fact that while cannon themselves represented an enormous change in warfare and continued to develop over the period, the challenge that cannon posed remained fairly constant. By contrast, the emergence of exploding shells, high explosives and the industrial power necessary to produce these in staggeringly large quantities posed a radically different, new set of challenges which, as we’ll see, fortress design struggled to adapt to. Because the changes in this period (c. 1870 to present) are very rapid this essay is going to be even more of a high-elevation overview, but I hope this will at least give a general sense of the shape of things.

As always, if you like what you are reading here, please share it; if you really like it, you can support me on Patreon. And if you want updates whenever a new post appears, you can click below for email updates or follow me on twitter (@BretDevereaux) for updates as to new posts as well as my occasional ancient history, foreign policy or military history musings.

Ring Fortresses

Despite this tremendous increase in available firepower, many of the same strategic concerns motivated fortress construction in the decades prior to the First World War. The experience of the Franco-Prussian War had demonstrated to everyone that quick offensives supported by railroads could potentially overrun even relatively large countries (like France!) faster than those countries could actually mobilize their armies. Consequently, countries that worried they’d be slow to mobilize (Russia, Austria) or that they’d need to wait for other powers to come to their rescue (Belgium), looked to use permanent fortresses as a way of delaying offensives, slowing down that initial rush until they could have their own armies ready for a decisive field engagement (recall that no one, at this point, has envisaged entire fronts of a war settling down to trench warfare; the assumption is that there will be a decisive series of pitched battles, albeit on a massive scale). On the other hand, fortresses could also be intended to shape offensives, closing off certain lines of attack (the fancy military term we’ll come back to here is ‘canalize’); France engaged in this kind of fortress construction prior to both world wars.

The solution that European fortress designers came to were ‘fortress rings’ composed of ‘ring forts,’ the latter term a bit confusing because a ‘ring fort’ or ‘ring castle’ was an older style of keepless-castle during the Middle Ages. Here, as far as I can tell, the forts are termed ‘ring forts’ not because of their shape, but because they are arranged in a mutually supporting ring around a key defensive point. That point was typically a city, but what the forts are defending is no longer the population or administrative center, but the infrastructure center, since cities tended to be where roads and railroads came together which made cities ideal blocking positions to deny the enemy access to that transportation system (which modern armies increasingly relied on to move with any speed and to manage their logistics). That in turn matters because armies are no longer drawing most of their supplies from local foraging operations, but rather are now primarily being supplied by railway lines reaching back to depots and factories in the home country, in part because by this point ammunition requirements (particularly artillery) by mass and volume were now beginning to eclipse demand for things like food which might be acquired locally. Artillery shells, alas, are not regularly grown by farmers out in the countryside. Thus, prohibiting an enemy army the ability to use the railroad and roadway junctions within a major population center could dramatically limit the ability of that army to move beyond the city. In practice then, this is an effort to achieve the same kind of operational blocking effect that castles with cavalry detachments did.

We can take the Fortified Position of Liège (PFL or position fortifiée de Liège) as a good example of the system. Built beginning at the tail end of the 1880s, the fortress ring was made up initially of 12 forts of varying sizes, arranged in a very rough circle with a radius of around 5 miles, centered on Liège. The dramatic increase in the range and power of artillery meant that, despite these forts being widely spaced, they could support each other with heavy artillery fire and ideally also use their artillery to support a screen of infantry which, in a war, would be dug in in the roughly 1-3 mile wide intervals between the forts. In theory then, an attacker would be unable to isolate the forts from each other; pushing through ought to require neutralizing the forts one by one (though it doesn’t, in the event).

Via Wikipedia, a map of the PFL. Our focus is on the blue forts which were built prior to the First World War.
This map also quite brilliantly shows how the city itself sits at the nexus of nearly all of the local transportation routes, such that an army looking to traverse the area is going to need to take control of the city merely to use the road and railway network to ensure supplies.

The design of the forts themselves were also responding to increasing firepower. The development of stable high explosives in the 1880s had triggered what became known as the ‘torpedo-shell crisis’ – torpedo-shell being the term at the time for artillery shells with a modern pointed cylinder shape and a high-explosive detonating charge inside. Such torpedo-shells (again, today we’d just call them ‘shells’) could punch into the earth under a fort and then explode with tremendous force, heaving the earth (and the poor defenders) upward. Such shells could easily obliterate even earthwork backed walls (like the old trace Italianne walls) or the above-ground brick casemates of older polygonal forts and they could do so from tremendous range, because unlike older siege guns, the striking power was in the explosive charge of the shell, not its raw velocity, meaning that it was no longer necessary to situate a cannon close to a fort for a direct-fire bombardment to breach the walls. This new danger forced the redesign of many older forts (for instance the Séré de Rivières system in France of which the fortress-system at Verdun was a part) and many of the fortresses of WWI were in the expensive process of modernization when the war started. The PFL, however, was begun relatively late (1888) and took these problems into consideration.

First, the forts, rather than being elevated above the surrounding terrain, were essentially submerged, built entirely in concrete below the level of the glacis so that enemy artillery couldn’t fire directly at the fort itself. The core armament of each fort, the heavy artillery, would be carried in rotating thick steel turrets (much like the guns of a contemporary pre-dreadnought), made to be nearly flush with the ground (to make it very difficult to land a direct hit on them with direct-fire artillery); ammunition was stored deep underground beneath these turrets and brought up, again, in much the same was as the main armament of a pre-dreadnought. The guns in question were quite large, with the larger forts having two 15cm guns, four 12cm guns, and two 21cm guns split between five turrets. To defend the heavy guns from infantry assault, each fort featured both smaller artillery (5.7cm) for direct-fire anti-infantry support. As with the older polygonal forts, the (dry) moat (here technically called the ‘gorges’) provided infantry fighting positions using the glacis as the edge of a trench, while at the same time the fort’s structures set down within the moat had firing holes and so could function more or less like caponiers, though as the firepower available to the defenders increased, it seems clear that concern about close-in-fighting in the gorge was lessened. Attempting to storm one of these forts on foot without first demolishing it through artillery must have seemed a mad gesture (and indeed, Russian efforts to storm somewhat older forts at Przemyśl this way failed miserably).

Via Wikipedia, a diagram of one of a few patterns for the forts designed by Henri Alexis Brialmont for the PFL. The larger artillery pieces are operated from the turrets (the black circles), while the actual living spaces for the garrison are actually in the ‘counterscarp’ – the outer wall of the fort’s gorge – which proved to be utterly untenable under bombardment.
Via Wikipedia, a sideview diagram of an idealized form of one of these forts; less accurate than the above in terms of actual shape, it expresses the vertical nature of the forts much better, with the ammunition magazines buried deep beneath the fortified gun turrets and the rest of the fort organized around that.

Moving to other ring fortress systems, where was substantial variation in the design of individual forts, often being shaped by the terrain and also because of the rapid pace with which they had to be redesigned and modernized to resist ever more powerful artillery. The Austrian fortress system ringing Przemyśl was less modern than that at Liège, but similarly based on the idea of mutually supporting artillery positions set in a ring, with the forts once again constructed as galleries dug down into the faces of the gorges, although much of Przemyśl’s older construction was in brick rather than concrete. The forts of the French defense system at Verdun were, in their 1914 form, closer to the design at Liège, although the French forts had been covered over with substantial amounts of reinforced concrete to fortify them against those torpedo-shells.

Via Wikipedia, a diagram of the fortress at Przemyśl. Though it had more fortified positions than the PFL, most of these were older and hadn’t been modernized. Nevertheless, the Russian army was unable to carry the fortress by storm and had to settle in for a siege, which it won in March, 1915.

These fortresses were impressive, they were often center-pieces in the defense strategy of their countries and they mostly failed. The forts at Liège (and similar positions at Namur and Antwerp) had been designed to withstand 21cm artillery with the firepower of 1886; the German Army, knowing this, had specially designed 42cm ‘Gamma Guns‘ (and the famous 42cm howitzer, ‘Big Bertha‘) with the specific aim of being able to breach the fortresses of the day. At Liège , the other problem was that a barrage, even one that might not destroy the fort, rapidly rendered the fort spaces unlivable due to ventilation, water and sanitation issues caused by bombardment. This problem was intensified by design issues, namely that the living spaces in the fort had been pushed into the ‘counterscarp’ (the outer edge of the gorge wall), which wasn’t exactly the most pleasant or safest place to be when the fort was being shelled.

In the end, while the Belgian fortifications did buy some time for Belgian, British and French mobilization, it was far less than the month that had been planned (though far longer than the two days that the often unrealistic German timetables of the Schlieffen Plan had allocated). Invested on the 5th of August, the German army had punched through the thin defenses connecting the forts and taken Liège on the 7th and were then able to isolate and reduce the forts one by one, with the last surrendering on the 16th. The subsequent Siege of Namur went even more smoothly for the Germans, with the fortress system falling in just five days and the garrison of some 35,000 defenders being neutralized at the cost of only 900 German casualties.

Via Wikipedia, a picture, taken in 1914, of one of the destroyed fort turrets at Liège. The forts proved unable to resist the firepower of German guns which had been specifically designed to destroy the known defenses.

Ironically, the collapse of the Liège forts convinced French high command that these fixed fortifications were unlikely to be held against determined German assault, leading to the progressive downgrading of the Verdun fortifications (their artillery was mostly removed for use amidst the trenches), despite the fact that the modernization efforts on the Verdun forts actually did render them relatively more resilient against artillery barrage, something the French would themselves find out when trying to retake forts Douaumont and Vaux after both were captured by the Germans in the opening of the Battle of Verdun. Fort Douaumont was captured by a small German reconnaissance party just four days into the battle, its defenders having been forced deep into the fort by the barrage and being caught unawares. Fort Vaux stood off the Germans better, holding out until June, but its defense was seriously hampered by the removal early in 1915 of nearly all of its artillery, leaving only a pair of 75mm guns in the turret. The French would retake both forts only after stunningly heavily bombardment; when defended, the forts seemed to work just fine.

In the more open environment of operations in the East, the other limitation with these fortresses became apparent: blockade. The fortress system at Przemyśl was nowhere near as modern as those at Liège or Verdun, but then the Russian army which laid siege to it in 1914 lacked the heavy fortress-busting artillery of the Germans. Efforts by the Russian Army to actually assault the forts or blast them initially failed. A first effort to take the city started in September, 1914 failed when the fortress was relieved by advancing Austrian armies on the 9th of October; but by the end of the month the momentum shifted and the Russians were back. Efforts to relieve the fortress by attacking through the Carpathians in winter – a series of catastrophic military mistakes by Conrad von Hötzendorf so bad that they nearly match the blunders of Luigi Cadorna – actually inflicted more losses on the Austrians (around seven times more losses) than were in the fortress itself, which fell in March, with more losses (including the surrender of the surviving 117,000 defenders) on the defender’s side than the attackers (not counting the Carpathian offensives), despite the defenders, in theory, having a fortress. Nevertheless, the Russians took severe losses in ill-conceived efforts to storm the forts with insufficient artillery before simply settling down for a siege and waiting the defenders out.

The Russians fared much worse in trying to hold fortresses against the Germans. Kicking off the Gorlice–Tarnów offensive on May 2, 1915, the Germans reached Przemyśl – now in Russian hands – on May 16, and captured the city on June 5. As Russian armies fell back in what became known as the ‘Great Retreat,’ fortresses which ought to have been able to buy valuable time instead became expensive traps. The Russian fortress at Kaunas had a garrison of 90,000 soldiers, while a large part of the Russian First Army, another 90,000 men, threw itself into the Novogeorgievsk fortress. The Germans took both handily; Kaunas was besieged on August 8th and the complex had fallen by the 19th while Novogeorgievsk was invested on the 10th and fell on the 20th. The loss of both, in particular complete with their shells and guns (2,900 artillery pieces and well over a million shells between them) was a disaster. The difference here was more ample German artillery – heavy artillery in particular – which made battering the defenses down practical.

Russian artillery captured at Novo Georgievsk.jpg
Via Wikipedia, captured Russian guns being removed from the fortress at Novogeorgievsk in 1915. Artillery and shell shortages had been a huge problem for the Russian army already, so to lose so much of both without inflicting much of any damage on the enemy was a catastrophe.

But of course that statement was a damning one: a fortress which is only useful against enemies with substandard artillery and crippling shell shortages isn’t much of a fortress at all. In practice, while forts could resist assault for a time (as at Vaux or Przemyśl), in the face of the firepower now available to the attacker and vast industrial armies which could easily detach armies of tens or even low hundreds of thousands of men to bottle up and besiege a city without even pausing an offensive, the sort of pre-built static defenses that had defined European warfare since the initial development of the trace italienne were now quite clearly on the wane. But of course this is World War I, so that doesn’t mean fortifications are on the wane.

Ditch Apotheosis

We’ve already discussed the trench stalemate that emerged on the Western and Italian fronts in WWI and to a lesser extent in many of theaters as well, so we don’t need to belabor the point here. What I want to note here are some of the ways that the trench networks that emerged by the end of the war reflected responses to the problem of heavy artillery, penetrating high explosive shells and firepower generally which had largely doomed the great ring fortresses of the pre-war era.

In particular, the systems of trench fortifications eventually embraced defense that was in depth and elastic. We’ve seen defense-in-depth as a concept quite a bit here, but again the fundamental notion of defense in depth is that because of friction (in the Clausewitzian sense, drink!) an attacker’s attack is never better organized nor stronger than at the moment it steps off; by forcing the enemy to engage a series of obstacles and disruptions, the defender turns that friction into their ally as the attack slowly breaks down due to confusion, casualties and random chance. While the pre-war ring fortresses in theory, by having multiple forts, embraced a defense in depth, in practice neutralizing just a handful of the forts in the ring compromised the entire position (in the case of Liège, the Germans actually compromised the whole fortress system without taking any of the forts; they punched through to the city first and then reduced the forts individually afterwards). An attacker could reduce the forts one-by-one, with pauses in between, or reduce just a few and then push through and take the rest from the rear. Consequently, the attackers could prepare carefully (in some cases, prepare for years since the fort’s layouts and defenses were well known) and unleash their full, un-frictioned attack on each fort one by one. It wouldn’t be inaccurate to say that the ring fortresses defeated themselves in detail.

By contrast, trench systems were deployed in depth. German doctrine was the most clear and sophisticated here and it eventually called for a standard three-line system (see below) with forward fighting positions which were to be abandoned, a main battle position, and then a third line with reserves which would, once the enemy attack had run out, counter-attack. These could be supported in turn by strategic reserves rushed along rail and road lines to the battle. British and French positions weren’t always so deep or well-organized, but they too were generally set in multiple mutually supporting lines. The problem, as we’ve discussed, wasn’t taking the forward positions, which was typically accomplished, but reaching the second or third lines with enough momentum to carry out into the open country, which was almost never achieved.

A fairly crude diagram of German late war trench defense doctrine, showing three lines of fighting positions. The folks over at The Great War channel made a much better diagram for the Hindenburg Line (1917) as a poster, which one apparently can still buy on their EU shop. I have one, its great.

Defense in depth was also a response to the massive increase in firepower, particularly artillery firepower. By holding the front line with only a light force and keeping most of the troops in reserves to the rear, the bulk of the defending force could be kept out of the enemy’s initial, devastating barrage, safely out of range of all but the largest enemy guns. In essence, the thin defense of the first line of trenches functioned as a screening force, shielding the fresh troops to the rear from enemy artillery, both by physically keeping the artillery back and out of range but also by helping to shield those rear areas from observation. The importance of observation swiftly becomes crucial when artillery are capable of devastating fires at over-the-horizon targets. While we often imagine artillery just indiscriminately blowing up everything in a large area, in practice artillery is only really effective if it can put shells on target and even a notionally dense front line is still mostly empty ground. Consequently, if you can deny enemy observation, you can dramatically reduce the effectiveness of their artillery (there’s also the related issue of shot adjustment; complex atmospheric conditions means that adjusting after each shot is more accurate than firing blind from range tables and formulas). As we’ll see, that focus on not being observed as a key component of defending against modern firepower is only going to become more prominent.

Finally, this is an elastic defense.1 Elastic defense was hardly a new concept; Roman frontier defenses, as we noted, were elastic on an operational level and the many layers of large, complex trace italienne forts were meant to be elastic in a tactical sense. Elastic defense is a much more active model of defense, wherein the defender does not merely receive and parry attacks but makes their own spoiling attacks and counter-attacks, both to regain defensive positions ceded to the enemy but also to disrupt enemy plans.

While in theory the pre-war ring fortresses had positions from which the defenders might launch sallies out, in practice those routes were generally cleared of obstructions and so would be swept by enemy fire too; the crushing power of modern industrial firepower made such direct sallies into no man’s land ill-advised. Moreover, secondary positions from which the defenders might launch assaults to dynamically retake defensive positions were typically missing; at best one might hope the garrison of the next fort down the line might try something, but in the event that fort, being fully exposed itself on the front line, would be in no position to lend aid to its neighbors, much less stage its own fortress assault. Instead, the defense of a ring fortress was typically rigid; forts once taken were not generally retaken; the clear exceptions at Verdun having more to do with those forts having been integrated into a trench network than being features of the fort’s original design.2

By contrast, the trenchlines of the Western front were always intended as both defensive positions and offensive stepping off points such that retaining the ability to ‘go mobile’ was crucial. Artillery might be dug in, with added overhead cover and earthwork berms to protect against enemy shelling, but the guns in question were still intended to be mobile and able to be redeployed from one sector to another (and indeed, just about every assault plan assumed that the attacker’s guns would eventually need to be moved forward to cover the advance, although actually doing this over the morass of the trench stalemate was often nearly impossible), while infantry was expected (by the end of the war) to dynamically shift from defense to counter-attacks as the opportunities arose.

As the war wore on, the importance of counter-attacks forced alterations in trench designs, leading to the main fighting positions being moved into the second trench line such that the intended response to an attack was to attrition the enemy before allowing them to take the first trench line, then stopping their attack at the second line – the main battle position – before reserves in a third set of positions would move up through pre-cleared communications trenches to launch a counter-attack to retake the forward positions. Since the enemy was likely to still be trying to push forward, those counter-attacks were also supposed to be flexible, shifting from defense to counter-attack as conditions changed (see the diagram above).

Thus it is a strange irony that the war which saw the beginning of the decline of fortress cities (not withstanding France’s effort with the Maginot line) became a war known for the strength of its fixed defenses, which remained virtually unmoved after four years of carnage.

Field Fortifications in the Modern System

As we’ve discussed at some length, the last years of World War I and much of the military developments of the interwar years were directed towards avoiding the brutal trench stalemate; the end result of these developments was what Stephen Biddle describes as the ‘modern system.’ Incorporating developments in mobile warfare (that is tanks, combined with motorized or mechanized infantry), air power (particularly close air support) and infiltration infantry tactics (also known as ‘Hutier‘ tactics), the modern system aims to use mobility, cover and concealment to avoid the overwhelming firepower of modern industrialized war in order to avoid the trench stalemate and retain a war of maneuver. Not every modern country’s military methods embrace the modern system, because it is expensive and difficult, but the modern system finds expression in a range of industrial military doctrines, particularly those of the most capable militaries.

This continued development of firepower has led quite a lot of observers to thus assume that the offensive is now all-dominant and that defenses, including fortifications, are a relic of the past (an idea often expressed through Patton’s misquoted contempt of the Maginot Line, presenting it as the both the last great fortification and also the marker of fortification’s failure). But as ADP 3-90 Offense and Defense (2019)3 notes, “While the offense is more decisive, the defense is usually stronger,” (4-1) a statement that is a near direct quote of Clausewitz, “the defensive form of warfare is intrinsically stronger than the offensive…yet has a negative object” (6.1.2, p. 358 in the Howard and Paret translation; drink!).

Playing the Clausewitz drinking game with US Army field manuals would be, by the way, a pretty good way to get really hammered really fast, but you’d need to know your Clausewitz, because the borrowings are typically unmarked.

That said, the defense envisaged in ADP 3-90 is not a static defense. Instead, the manual notes “a defending force does not wait passively to be attacked” but “seeks ways of attriting and weakening enemy forces.” Maneuver and concentration even on the defense is stressed, as is a willingness to give ground in order to either buy time for maneuver or wear down enemy forces. In short then the defensive stance that is envisaged here is an elastic defense, but one that stresses even more flexibility and speed, which fits with the modren system, especially since the mobility demanded here is in part to retain concealment since once a force has been engaged, its position is known and so it must move in order to re-conceal itself. And of course the goal of the defensive, here as in Clausewitz (drink!) is always to get back on the offensive.

So what place do fortifications have in this vision? Quite a bit, actually, but the mobility and flexibility of the doctrine lead to an emphasis on far more temporary defensive positions, held for a given purpose and given up as soon as that purpose has passed. The ATP 3-21.8 Infantry Platoon and Squad (2016) field manual lays out the basics for small units. “Preparations” – which include both utilizing natural obstacles and enhancing those obstacles to aid in a defense “end only when the defenders retrograde or begin to fight” and that “preparations in-depth continues, even as the close fight behinds.” Strongpoints designed to block or canalize enemies are important; the manual notes such a position typically takes roughly a day for an engineer detachment of equivalent size to construct (with ATP 3-37.34/MCWP 3-17.6 Survivability Operations laying out the ways that existing structures and terrain can be reinforced). The emphasis in positioning, as with late WWI-trenches is in channeling (‘canalizing’) enemy attacks into areas with overlapping fields of fire or exposed to indirect fires (that is, artillery). But the modern system emphasis on maneuver comes out clearly as well. The manual twice lays out expected phases or sequencing of the steps for a defensive action (in 3-47 and 3-70), in both cases ending in swift maneuver and counterattack. You let them hit you, defeat their advance, maneuver against them, and then strike before they can attack again, getting back to mobile warfare as quickly as possible.

From ATP 3-21.8 Infantry Platoon and Squad, a diagram of a defensive position to stop a two-pronged enemy attack. Note the use of obstacles (rectangles with three V-shapes) to channel the enemy advance (enemy is marked as EA RED and EA BLUE) away from the flanks and into overlapping fields of fire. Crosses (with AB numbers) are planned indirect fires (read: artillery or mortars).

Fighting positions (discussed in ATP 3-37.34/MCWP 3-17.6 Survivability Operations, chapter 4) are fortified, but they aren’t fortresses. At the beginning are ‘hasty positions’ – using what the local terrain and structure offers, combined with basic dig in to protect against direct fire (though the field manual notes such quickly dug positions offer “no overhead protection” – the concern here being the modern version of those bursting shrapnel shells). “Deliberate positions” follow with more time (once again “continued improvements are made…during the period of occupation,” so the longer you are there, the more works you put up), with an emphasis on concealment (using camouflage, or existing structures), overhead cover, and opening protected routes connecting positions. Some of the positions described imply several days of preparation, but not too much more because of course the longer a fighting position is occupied, the more likely it is to be spotted and hit with overwhelming enemy firepower. The doctrine here largely does not address situations where formations are in contact with the enemy for months at a time because in this doctrine that isn’t supposed to happen (whereas in WWI that was the exact condition in which the trenchlines emerged).

Also from ATP 3-21.8 Infantry Platoon and Squad, a diagram showing how a fighting position for an armored vehicle can be developed depending on the time available, with a natural depression used initially, but then dug out to make both hull and turret defilade positions, plus camouflage netting used to disguise the vehicles presence from aerial observation.

I should note that all of these manuals also have considerations the impact of tactical nuclear weapons on these defenses Defense in depth is recommended in ADP 3-90 expressly when there is a threat of weapons of mass destruction (much the way it was used to limit exposure to massive conventional artillery barrages in WWI) and ATP 3-37.34 assesses fighting positions based on the protection they provide, inter alia, in the event that nuclear weapons are used. In essence the potential of the use of nuclear weapons takes the firepower/concentration problem and dials it to 11. On the one hand, the defender, like an attacker, wants to be concentrated so as to deliver maximum force at the point of decision (in this case, at the point where the enemy attacks), but also wants to be dispersed so as to avoid losing the entire unit to the devastating impact of modern firepower. Concealment and mobility are presented as the solutions: concentrations can be maintained only for the decisive moment and then an immediate return to maneuver to deny as much as possible that concentrated, target-rich position for the enemy to deploy their massive firepower, be it conventional or nuclear, on.

Meanwhile, US Army doctrine envisages more extensive fortifications for permanent base camps and forward operating bases (FOBs); these aren’t (in theory) frontline fighting positions, but could come under attack (especially indirect fires or guerilla attack) and so have to be fortified against that kind of attack. The basics are covered in ATP 3-37.10/MCRP 3-40D.13 Base Camps (2017), which in turn frequently refers back to ATP 3-37.34 for the particulars of fortification design. Such bases are to be fortified, typically with a single-layer perimeter (that is, a a non-concentric defense, but that defense typically has multiple elements) in a rectangular shape supplemented with internal barriers. The outer barrier generally consists of a combination of barbed wire, a physical barrier (that is, a wall) and earthworks (typically ditches, but also berms). The wall can be made out of many kinds of materials, but as far as I can tell, the HESCO barrier, made up of a wire-mesh container which can be rapidly filled with earth and set in rows and stacked to create a barrier, is the most common. Interestingly, this is an old technology come back: the venerable early modern gabion in a modern, industrialized form. As with the gabion, the advantage of HESCO barriers is that it ‘catches’ projectiles rather than either shattering or causing the projectiles to ricochet. The field manuals also note the use of similar barriers within the base to catch shrapnel from enemy artillery or mortar fires (see the image below on how you can ‘compartmentalize’ shrapnel this way). So a complete perimeter might consist of a ditch, then a line of HESCO barriers, topped with wire to prevent scaling, with the interior base then subdivided by more earthworks or HESCO barriers to reduce the impact of shrapnel.

Via Wikipedia, a German base inside Camp Marmal in Afghanistan in 2006, showing the heavy use of HESCO barriers both to protect the internal structures from direct fire but also to control shrapnel from mortar shells that land inside the base.

That basic perimeter is then reinforced with fighting and observation positions. Towers sit inside the perimeter (often on the corners) but flush with the outer wall to maximize observation while still benefiting from the protection of the perimeter wall; they can double as (and also be supplemented by) dedicated fighting positions on the perimeter. The gate – or ‘entry control point’ (ECP) is also reinforced and guarded – access points are to be designed so that vehicles can’t approach the perimeter in a direct, perpendicular line, but have to turn into the base, using “sharp turns and serpentine layouts” to mitigate high speed vehicle attacks and generally frustrate enemy efforts to access the gate.

From ATP 3-37.10/MCRP 3-40D.13 Base Camps, a diagram of observation positions on a wall noting how towers placed flush with the wall avoid deadzones while still benefiting from the protection of the barrier (which could be made of many materials, including HESCO barriers, but also stone, earthwork, concrete, etc.).

And at this point I hope this combination of features is familiar: a rectangular shaped outer earthwork supplemented with an anti-climb device on the top, ditches to frustrate approach and observation towers flush with the wall on the inside of the corners, with a gate reinforced with a serpentine layout…because that description could also describe a Roman marching camp. And for much the same reasons, because while the types of weapons have changed, the basic threats the design aims to avoid are the same: the goal is to maximize observation while slowing down enemy access to that the army inside the camp has time to mount an effective pitched-battle response. Both designs also show a significant concern with the possibility of enemies seeking to fire into the camp from a distance; both often place an interval between outer barriers and the internal structures to mitigate the risk of weapons thrown into the camp (be they grenades or javelins). I don’t know that US Army planners aimed to reconstruct a modernized version of the Roman playing-card fort, but the concerns of force protection for an army on the move (and then an army settling down to try to establish durable control and engage in COIN) remain largely the same and so lead to similar designs.

Also from ATP 3-37.10/MCRP 3-40D.13 Base Camps a standard ‘rectangular box’ design for a base camp. Naturally, the logistics concerns demand significant changes from the Roman ‘playing card’ fort (particularly the need for vehicle access and storage) but the overall principles remain strikingly similar.

One wonders how central these sorts of fortifications will be in the history of the American limes; will archaeologists marvel at the striking similarities in design between American FOBs located many hundreds of miles from each other? But in any case, modern warfare has not removed the need or value of field fortifications, although within the modern system, the emphasis on mobility remains very strong. But is there a future for fortifications beyond the modern system?

A2/AD and the Future of Defense

Predictions are hard, especially about the future and not generally the province of historians. Nevertheless, defensive technology has not stood still and it could be argued that several new technologies may shift the balance back from the apparent supremacy of the offense and thus bring back a form of fortifications, albeit of a kind that won’t look very much like the castles and concrete pillboxes of old. Still, such technologies aim to accomplish the same goals, raising the cost of the offensive unacceptably high and channeling attacks that do occur and so we might well term their use a form of fortification.

Now defense and fortifications are not necessarily identical. Nuclear deterrence is, after all, a form of defense mostly4 without fortifications: a defense through almost pure, offensive retaliatory power. That said, missile defense aimed at interdicting ICBMs opens the possibility of fortifying an area against nuclear strike using ICBMs. My sense is that public perception of missile defense begins with SDI (the ‘Strategic Defense Initiative’ also known as ‘star wars’) in the 1980s, but the idea of a fortification system designed to prevent a nuclear strike goes back at least as far as the Nike-Hercules missile, developed in 1958 and the later Nike-Zeus, both of which could in theory intercept ballistic missiles (with their own nuclear payload), though never with sufficient reliability. As a personal aside, when I was a kid, very near my elementary school was ‘Nike Park’ so named because it was one of the 13 Nike Missile sites positioned around Washington DC to provide the capital with a defense system; decommissioned in 1962 it was turned into a park with the old radar dome being left as a cool echo chamber.

While SDI didn’t fulfill expectations, the 90s and aughts saw leaps to where missile defense systems now appear to be feasible and effective and a number of different countries have developed them. The United States has the Ground-Based Midcourse Defense, administered by the Missile Defense Agency, which manages a $9.18bn budget. While that reflects a significant investment in absolute dollar terms, compared to military spending more broadly (both in the USA and globally) it reflects a belief that the probability of a nuclear exchange is low. Consider that the annual cost of the Missile Defense Agency is just 0.04% of the United States’ GDP; for comparison, the initial construction of the Liège forts (the PFL) alone cost more as a percentage of GDP (around 0.065% if I have done my currency conversions right; alas I have the GDP of Belgium in 1880 in 1880Int$ but the cost of the forts in Belgian francs) to build and Belgium of course built equally expensive fortress rings around Namur and Antwerp and then had to supply and man these fortifications.5 The problem, at least for now, with ballistic missile defense is that the two leading nuclear arsenals, those of the United States and the Russian Federation are so large that they could simply overwhelm current defense systems, but as a defense against smaller nuclear powers, the possibility of fortifying a country with anti-ICBM systems now appears, if not realized, at least within the realm of possibility, especially if a rising sense of nuclear threat caused a radical increase in the level of investment in such systems.

Meanwhile, conventional warfare and defense hasn’t gone away. While the tradition fortresses of previous eras have largely been discarded in the post-WW2 era, the so-called Revolution in Military Affairs has made stand-off precision weaponry a possibility, extending the range from a border or coastline that a defender can seek to exclude enemy forces with the threat of precision strikes delivered by cruise missile or drone. These sorts of systems fit under the headline of A2/AD (anti-access, area denial) weapons because their goal is to prevent an enemy from moving through or operating in a specific area (though the term can also mean older forms of area denial weapons like land mines or even caltrops, the meaning of A2/AD in policy circles these days almost exclusively refers to this very modern meaning of the term). A2/AD systems allow a country which might not actually be in a position to directly contest a battlespace (because its army or navy is not confident of direct victory in that space) to nevertheless deny it to an enemy by ensuring that the area is covered by enough precision systems that the attacker would take unacceptable losses attempting to operate in the space, which can in turn mean denying the enemy the ability to actually deliver forces to one’s own country as a form of defense. The discussion of these sorts of systems has been most prominent in terms of China’s sea-directed A2/AD system, but such stand-off A2/AD systems look to be a durable component of national defense, at least for countries wealthy enough to afford large numbers of the expensive and sophisticated systems.

The emergence of the related ‘porcupine’ strategy, particularly in the context of the country6 of Taiwan, which aims to combine both the potential of A2/AD systems with the potential intractability of a motivated insurgency to raise the cost of attacking even a small (but wealthy, because some of these systems aren’t cheap!) country to unacceptable high levels, presents another avenue for future fortification. The core of a porcupine strategy, as laid out so far, is the use of both anti-land cruise missiles, anti-air and anti-ship missiles based in expendable and concealable platforms, either tractor trailers or small missile boats (which might, if small enough, be able to hide as fishing boats), combined with a final ground defense oriented more around irregular capabilities which would be difficult for an attacker to target with their own precision systems (that is, infantry with man-portable weapons, not expensive tanks and jets). Essentially, deter an attack by promising the enemy they’ll have to wade through lots of long-range precision strikes merely for the honor of getting to wage a long and difficult counter-insurgency campaign at the other end.

I think it is not at all unreasonable to see developments like contemporary A2/AD and the related ‘porcupine’ strategies as a return to fortifications of a type. Just as the industrial firepower of the late 1800s forced the single forts or fortress systems of the previous era to be broken up into the ring fortresses, so continued military development has demanded that defensive capabilities continue to be broken up, spread out and most importantly concealed in order to be effective. But a country which invests heavily in mobile, concealable A2/AD assets in order to deter attack is engaging in fortification of a kind, building physical infrastructure for the purpose of deterring or stopping an enemy attack.

So are “fixed fortifications…a monument to the stupidity of man?” Well, once again, because I hate misquotes, I feel the need to note that what George S. Patton actually said was, “this is a first class case of man’s monument to stupidity” about the Maginot line. But more broadly, the answer here is clearly ‘no.’ Fortifications made sense historically. The current popular conception that ‘fixed’ fortifications are useless is a consequence of a shift towards the offensive, particularly during the Second World War, but the balance between the offensive and the defensive is ever-shifting and never stable and may even now be shifting back towards the defensive due to the relative vulnerability of conventional offensive military forces to modern precision weapons (and the relative difficulty of using those same systems effectively against local insurgent forces).

In the end, the continued relevance of fortifications comes from the continued strategic imperatives: for weaker powers to find ways to deter attackers or at least shift the balance of strength so that a weaker power might win and for stronger powers to force a conflict to occur in favorable conditions. So long as there are polities operating under those conditions, they are going to try to find ways to fulfill those objectives. Not all of those methods will be fortifications, mind you – we haven’t much discussed insurgency here, but it can be seen as a non-fortification defensive strategy when adopted at the state-level as a planned fallback for when conventional arms fail – but some of them will be. The methods change, but the enduring strategic imperative means that fortifications, in one form or another, are also likely to endure – though we should also be aware that just as the forms of fortifications have changed radically in the past, they are likely to change radically in the future.

Happy New Year everyone!

  1. Elastic defense and defense-in-depth are often treated as synonymous, but I am actually trying to build a distinction here between a defense that is designed to fail in sequence and utilize enemy friction against them (defense in depth) and a defense that assumes quite a bit of active maneuver, envisaging both retreats and advances within an overall defensive framework. In this sense an elastic defense is a sub-type of a defense in depth which expects to transition flexibly between its various layers, rather than merely falling back to each in turn; under this rubric all elastic defense is in depth, but not all defense in depth is elastic. For instance a defense-in-depth of a castle that is non-elastic is one that expects to first defend at the outer wall, then an inner wall, and then the keep, and hopes that by the time the assault reaches the keep’s door, the attacker is too spent to continue. By contrast an elastic defense of the same castle might envisage sallies beyond the wall, or sorties to retake parts of the outer wall even after the enemy’s main assault has been launched.
  2. Some forts, particularly those in Belgium, were designed to be more vulnerable to assault from the rear, but this turned out to be a disastrous mistake as there was no ‘back-up’ system of defenses to prevent an attacker, having breached the ring, from then rolling up each fort in turn from its vulnerable rear, nor were they secondary positions designed to let the defender actually utilize this vulnerability, in stark contrast, for instance, to the hollow ravelins of trace italienne forts where the clear and obvious position from which to retake a ravelin was the bastion that sat behind it and could fire directly into its works. Alas at Liège the fort system had only half of this good idea which had perplexingly already been fully developed in the 16th century
  3. A United States Army field manual. I am going to lean quite a lot on US military field manuals here because a lot of US doctrine is declassified and available to the public, which gives a really clear view into how one of the world’s premier militaries functions and thinks about these things
  4. Of course individual nuclear installations – silos, command centers, and so on – are often heavily, even extravagantly fortified to render them resistant to nuclear attack, but that is merely to ensure the durability of the second strike retaliation. The actual target of a nuclear strike – cities, conventional military forces – cannot be sufficiently hardened to resist such an attack and typically no attempt is made to do so.
  5. Another comparison here might be the Manhattan Project, which is estimated to have cost around 0.8% of US GDP during its development.
  6. Yes, I said it.

156 thoughts on “Collections: Fortification, Part V: The Age of Industrial Firepower

  1. Interesting. Have you read about the Australian defense of Tobruk in the 2nd World War? IIRC, they relied on a fixed system of defensive positions, and a rolling system of random counter-attack and attrition of the DAK attacking forces – encapsulated in the claim that they “owned no-man’s land”.

    And one can consider General MacArthur’s policy during the Pacific war of taking only some of the islands held by the Imperial Japanese forces and interdicting the supplies of the other islands and letting them wither.

    As far as nuclear weapons go, I’d’ve thought that their near-unlimited power – particularly when on a Cold War level – makes them unusable as weapons. Thus to arm to a Cold War level with nuclear weapons to to fall prey to a Maginot Line paralysis on the military level, and a Parkinsonian form of (political) tremors on the political level. Not to forget, a paralyzing Interdependence with your enemy on the International level, where every policy having the slightest bearing on said enemy, is carefully checked for “resonance”. The USA and the USSR lost their respective independence in exchange for a paralyzing interdependence as a result of the Cold War and only regained it – partially – when the Cold War officially stopped in the 1990s.

    1. As an added bonus, public opinion forced the Japanese navy to use submarines to smuggle supplies to stranded soldiers. Not an effective tactical use.

    2. The Burma campaign notably featured elements of the British Eighth Army notionally on the defensive, but constantly tactically on the offensive breaking up enemy concentrations and doing everything possible to prevent an actual dangerous assault from coming together. The comparison used in the regimental history is apparently that of a boxer constantly throwing punches to keep his opponent from getting a good punch in.

    3. >Interesting. Have you read about the Australian
      >defense of Tobruk in the 2nd World War? IIRC,
      >they relied on a fixed system of defensive positions,
      >and a rolling system of random counter-attack and
      >attrition of the DAK attacking forces – encapsulated
      >in the claim that they “owned no-man’s land”.

      I suspect that strategy worked much better than it otherwise might because the Germans were at the fingertip limit of their logistical reach (Rommel was a good tactician but never had much regard for supplies; that paid off sometimes with bold mobile-warfare gambits, but tended to cost him if he stayed in contact with the enemy for any length of time). As discussed in the earlier parts of this series, it’s easier to disrupt an enemy’s siege of your fortress if they can’t get reliable supplies or (in post-gunpowder warfare) properly heavy equipment to blast you out of your defenses.

      >And one can consider General MacArthur’s
      >policy during the Pacific war of taking only
      >some of the islands held by the Imperial
      >Japanese forces and interdicting the supplies
      >of the other islands and letting them wither.

      This, in turn, worked because the US was in the process of acquiring air and naval superiority. American-held islands were effective bases for cutting off Japanese supplies; the reverse was increasingly not true after mid-1942. An American-held island could function as a launchpad (literally) for bombing operations that would cripple a Japanese-held island’s ability to pose a threat to anything out of visual range of that island; the reverse was increasingly not true.

      >As far as nuclear weapons go, I’d’ve thought
      >that their near-unlimited power – particularly
      >when on a Cold War level – makes them
      >unusable as weapons. Thus to arm to a Cold
      >War level with nuclear weapons to to fall prey
      >to a Maginot Line paralysis on the military level,
      >and a Parkinsonian form of (political) tremors
      >on the political level. Not to forget, a paralyzing

      Yes, and no. A small nuclear arsenal makes attacking you extremely costly (see the “porcupine” strategy in this article), because it only takes a few successful nuclear strikes to make the aggressor regret the war. Even if they still have a country afterwards, they’re spending trillions of dollars on reconstruction and burying millions of dead. This is a strategic-level extension of the same logic of making a fortress nearly unattackable by giving it so many interlocking fields of fire that the attacker cannot reasonably hope to storm the fortress and live.

      As to the “paralysis,” I would say ‘no.’ Heavier nuclear arsenals emerge as a way of ensuring that even a first strike under worst-case conditions cannot feasibly eliminate your ability to make an enemy bitterly regret having struck you at all. They are generally built up under the assumption that 80-90 percent of them will be destroyed before launch or never launched at all, or simply malfunction and fail to destroy the target.

      This is not paralysis, this is making sure you can take the enemy with you The “tremors” you describe, well, that kind of concern existed in all previous ages. No country ever acts in ways that will affect a foreign power without weighing the risks and costs of war with that power, unless that country’s leaders are fools. All that nuclear weapons do is greatly increase the risk of war and the immediacy with which the devastation can be inflicted. It doesn’t make you more interdependent, though it does make it more urgent that national leaders be able to communicate quickly and clearly!

      1. Well, an army marches on its belly, as Napolean is notoriously alleged to have said. So anything that keeps an army starving is going to render it useless. Rommel was the victim of an ineffective Luftwaffe assault on Malta, which cut off his supplies. The Tobruk Rats earned their reputation during his earlier stage, while he was still well-supplied. Basically, the Tobruk Rats made the siege of Tobruk as hard on the DAK as it was on them, which worked against the DAK, which like other parts of the Wehrmacht at that time, was geared to quick, overwhelming blows disrupting their opponents and driving them from the field.

        1. Rommel’s problem was not Malta (although that certainly did not help) but the trucking distance inside North Africa. Elsewhere, the Germans (and everyone else) relied on rail for long-distance support – trucks were for the last 300 kms. It’s 1250 kms from Tripoli to Tobruk. As van Creveld pointed out, you could unload supplies at Tripoli (short sea distance, good port capacity) and use most of the fuel hauling them to the front, or unload at Benghazi (longer sea distance, low port capacity) and get bombed by the RAF – losing tucks and ships. With Benghazi and Tobruk you can just get to Alamein, but you’re out of gas by then unless the enemy gives up.

    4. I’ve honestly had the same thought about nuclear weapons. They are so destructive as to be liable to destroy whatever one might hope to gain in war. There is no value in conquering an irradiated wasteland after all. So the offensive strategic use of nuclear weapons is pointless unless if genocide is the objective and not conquest. On a tactical level nukes have the same problem as chemical weapons. They deny the attacker access to the area they target due to radiation. So a breakthrough is functionally impossible after using nuclear weapons. Attempting to use a tactical nuke against enemy concentrations requires knowledge of where the enemy is concentrated. Modern system militaries never concentrate before the decisive point. Unfortunately, this means you will only get a concentration worth nuking after an engagement has begun, so your own troops are likely to be caught in the blast. And in cases where a modern system army is engaged with a more outdated force, then nukes are hardly necesary and their area denial is in fact a weakness.

      Defensively, nuclear weapons are just as useless. If they are used against invading enemies then you are nuking your own country. If they are used against the invaders homeland they invite retaliation in kind, but also dont solve the “currently being invaded” problem. Nuking the enemy homeland while their armies occupy yours isnt exactly a win. And on top of all this, any amount of nukes sufficient to overwhelm and destroy any great power is likely also enough to cause a nuclear winter and doom human civilization. So as long as the enemy isnt planning to kill everyone, the use of nukes is unacceptably risky. The greatest danger posed by nuclear weapons is that some lunatic death cult gets them and tries to blow up a city or two, not use by an actual country.

      1. Both the USA and the USSR concluded that it was not possible for either to “conquer” the other; even in the absence of resistance the sheer size of the territories involved and logistics on an intercontinental scale would make a classic occupation impossible. So each did envision genocide, or at least national extinction of the other. The point would have been simply to eliminate a chief rival, leaving the other as the hegemon of the planet.

        In the case of NATO v. Warsaw Pact, the Soviet Union war-gamed extensively with planned use of chemical weapons: troops in tanks and armored personnel carriers on the advance would be far less hampered than defending infantry. While the NATO forces considered overrun territory “lost” anyway and so planned to use tactical nukes especially where terrain would force the invaders to concentrate (Fulda Gap).

        West Germany (as it was then) indeed was less than happy about being on the front line of World War Three; that was part of the impetus for the development of enhanced radiation weapons (“neutron bombs”) that would at least try to minimize the destruction. And this was in the context of a “Flexible Response” doctrine: NATO would try to win conventionally if it could, only escalate to tactical nukes if central Europe was being completely overrun, and if that didn’t work use “theater” nuclear systems such as the Pershing II missile in an interdiction role as a last attempt to prevent a general strategic exchange.

        All of this was in pursuit of a combination of doctrine and technical means that would keep the use of nuclear weapons plausible enough to deter a conventional invasion of central Europe.

      2. Note that nuclear winter is vastly overstated and based on studies that were quite badly done.

        1. Two notes on nuclear warfare:

          (1) If the defender has enough resources to have tactical weapons, and the attacker has enough resources to be capable of attacking such a state, then once nuclear weapons are deployed there is no off ramp to strategic nuclear exchange. Even the weakest usable systems have 30-40 minutes to any target on the globe. And any weapon destroyed on the ground, as the Japanesse Imperial Navy learned at Midway, is useless.

          So, the moment any launch is detected the only possible decision is full strategic counter strike.

          In the specific Cold War scenario it’s even worse. There was assumed conventional suprioirity of the Soviets in Europe and it was NATO imagining the defensive use of Nukes, but the distance from Eastern and Centeral Europe to Moscow/Leningrad/Stalingrad/Minsk/Kiev/Odessa… etc is so short that the Russians would have to respond before they know the nature of the fight.

          And, to my understanding the US Army was never able to wargame a scenario with tactical exchange where the participants didn’t immediatly escalate to strategic exchange. This is also why arms control first focused on things like missle defense, cruise missile mounted nukes, and theatre nukes rather than strategic city killers. Things that could make the unwary politician or general believe that success was possible.

          (2) Nuclear winter may be overstated, but the simple explosive power and number of nuclear weapons is often understated. An American Ohio class submarine carries 24 missiles with 11 independently targetable war heads per missile. So, one Ohio class submarine can destroy the largest 264 cities in a given country.

          A similar weapons system used against the United States would destroy 50-60% of national population before accounting for secondary and tertiary effects. And this assumes no conservation of force. Because of not just cities like Minniapolis (46) / St. Paul (64) but also signifigant suburbs like Vancouver, WA (141) would be covered by the weapon that destroyed the primary target, Portland, OR (27). So, the point where someone is targeting 264 metro centers rather than individual cities you’re now destroying reltivaly small communities like Montpillier, VT and Aspen, CO… just because.

          (list of 200 largest US cities:

          1. Some countries can launch strategic nuclear weapons from nuclear submarines which are basically undetectable and can retaliate long after the first destructive strike on their homeland.

            I don’t know exactly how it goes in other countries, but France nuclear doctrine relies heavily – if not mostly – on those.

      3. > On a tactical level nukes have the same problem as chemical weapons. They deny the attacker access to the area they target due to radiation.

        This is not the case. In the 1950s both USSR and USA tested the concept of a breakthrough enabled by a tactical nuke by having mechanised infantry units advance right through after a nuclear test; it was confirmed that this is entirely viable – there is no area denial, the attacker can roll through pretty much right away. IIRC the soviet troops participating in the test had elevated cancer rates a decade after, but no immediate effects; and compared to the expectations of fatalities and injuries in such an offensive, the cancer impact is literally insignificant.

  2. Hardening cities against nuclear strike

    Occasionally, the notion is floated that the striking speed and extent of post-WW2 North American suburbanization was at least partly an intentional policy aimed at increasing the survivability of cities against a strike by the brute-force expedient of spreading cities, both population and economic activity, over a larger geographic footprint. Usually the idea is dismissed; the proximate factors unique to this time and area, most notably the actually-implemented policy to push motorways through cities (in contrast to e.g. London’s planned but not built Ringways system) are partially explainable as the result of unrelated causes.

    1. I recall a non-fiction article by Robert Heinlein from sometime in the 1950s or 1960s where he talks about how unworkable such spreading out would be, which means the idea was around then. Probably somewhere in his “Expanded Universe” collection.

        1. No, it is not connected to any of his fiction. Though he ended up writing a tiny story fragment inside it to illustrate his point. Poked around a bit, found it. It is a 1946 essay “The Last Days Of The United States”, unpublished until 1980 in his “Expanded Universe” collection.

          Turns out Baen Books still has it up as one of the sample chapters to their ebook version, despite the fact that they don’t seem to sell the ebook anymore (In my experience Baen’s website management err, ‘could be better.’) The terms he uses are “decentralization” and “dispersion.”

          You can see it here:

          1. Baen no longer sells post-publication-date ebooks as part of the deal that they made with Amazon to sell ebooks through Amazon instead.
            They *do* sell ebooks “pre-publication-date” (in bundles only) and if you buy them via Baen you can continue to download/view them there after they go off-market at Baen.
            (They also still have their “free library” – which books Amazon also carries for free)

      1. Yes! It’s his “The Last Days of the United States.” He’s talking about a lot more suburbanization than actually happened, pointing out that we can’t leave essential infrastructure in the old downtowns.

    2. The US’s suburbanization could not have been feasibly imagined by the planners responsible as an attempt to make cities more survivable to nuclear attack, because the blast radius of the bombs was growing faster than the physical footprint of the cities. By 1955 it had to be obvious to any competent nuclear war planner that everything within 5-10 kilometers of the point of impact of a nuclear device would burn, and that the number of devices thrown around in a serious nuclear war would number in the hundreds if not thousands, meaning that any given city of real importance would be hit with multiple bombs.

      Anyone who pushed for suburbanization as a form of security against nuclear attack was very, very foolish, or very, very ignorant.

    3. When the advent of ICBMs eliminated having enough time to conduct any effective evacuation, the potential of deep-sheltering to survive thermonuclear attacks was looked at. The Manhattan Shelter Study conducted by the Hudson Institute and the RAND corporation studied what it would take to save most of the inhabitants of Manhattan from thermonuclear destruction and declared it to be impossibly expensive. A faint echo of this idea was referenced in the movies “Beneath the Planet of the Apes” and “Battle for the Planet of the Apes”, in depicting irradiated survivors huddled far beneath the nuked ruins of Manhattan.

      1. I believe that as a neutral country protected by their banks, the Swiss weren’t trying to defend against a direct nuclear strike, but against a nuclear strike against one of their neighbors, so protecting a population living tens of km from the actual hit, possibly hundreds if it wasn’t too close to the border.

        I’m not sure whether it would have had a chance to work, anyway, but it’s a much easier target to aim for.

        1. I don’t know anything on the details of the particular claim, but the Swiss during the Cold War, and probably today, went to a lot of effort to be ready to actually be invaded. It makes sense to me that they were probably also trying to be ready for nuclear strikes on their territory. A huge part of Swiss strategic thinking was trying to make it extremely difficult and expensive for a bigger power to conquer them (and thus they could stay safely neutral and not worry about a hegemon arising in the way everyone else did because they would be able to negotiate acceptable terms with anyone who became dominant in the rest of Europe).

          1. The Swiss concept of the National Redoubt is as much a cultural project as it is an actual defense initiative, if we’re being completely honest here. It’s part of the national identity that Switzerland is neutral and also has never been invaded because the cost is not worth the benefit.

            Quietly not mentioned is that, during WWII, the government did actually allow right-of-way for German trains… And also that evidence of doing business with the Nazis was a national embarrassment for several of the country’s largest banks. I don’t recall if it actually destroyed any of the banks outright, but a few of them took significant prestige hits. Business being business, however, at least a couple of them are still reasonably successful today.

            I won’t say that the Swiss national defense strategy is bad, but I’d argue it is coming from a place where it has neither been tested nor particularly needed to have been tested. That’s why I’d call it more of a cultural project.

  3. Nit: “tradition fortresses” -> “traditional fortresses”.

    As far as radical changes in the efficacy of fortification systems of the future, I think the warfare of the Koprulu sector of the galaxy provides an interesting example. During the last days of the Confederacy, static defense was a powerful tactic: armies were strangely uncoordinated (despite having radio / psionic communications!) and photon cannons on the high ground or a line of sunken colonies were a powerful tactic, slowing enemy assaults at worst and sandbagging them entirely at best, unless outranged by some specialized long-range siege weaponry. But just a few years later, technology had advanced to render them far more restricted in use. Dropping random photon cannons around the map were essentially just donating free kills to coordinated enemy squads, which could easily move up and dispatch them without problem. Sure, maybe if you realize a roach all-in is coming, dropping a wave of Spine Crawlers is what you have to do to not die, but the loss in flexibility and the ability to threaten a counter-attack that simply building more mobile units was high. Bunkers were still decent, but largely because of improved recycling technology that allowed the resources used to build them to be reclaimed quickly after they were no longer required – hardly a stirring endorsement that they’re good because you can return them after they’re bad. Mostly because of easier coordination of armies, static defense simply became less valuable.

    1. “But we must remember that war involves an entire society. In the civilian sector, the development of multiple building select allowed larger armies to be created more efficiently, tipping the balance further against defensive structures. This improvement in terran and protoss production abilities may be the biggest factor in their increase in power relative to the swarm after Kerrigan’s apparent conquest…”

  4. Ring fortresses failing while trenches succeeded could make for a great image:
    “Who would win: An extremely expensive set of fortresses, built on the latest theory of siege warfare, or a few guys with shovels”

    I don’t know anything about modern weapons (or premodern weapons for that matter) but I heard news reports on hypersonic missiles (hypersonic meaning 5x the speed of sound). These missiles can travel so quickly that they could escape anti missile systems. Is this just some Russian technology to tell the west they can’t think they are safe behind a missile shield? Or does this mean that missile defence systems are not feasible?

    1. Even then Hypersonic missiles become more expensive and fragile than conventional missiles, potentially making lasers (especially given their continued development) a cost effective defensive tool.

    2. “Hypersonic missiles” are actually not notable for their speed – they are the same speed, by definition, as ICBM reentry vehicles. Rather, it is a (sloppy, non-technical) contraction of “missiles with hypersonic glide vehicles as payload” – that is, weapons that combine the speed of ICBMs and IRBMs with the controlled maneuvering of smaller and slower guided missiles or cruise missiles.

      These are the “A2/AD” systems Dr. Devereaux talks about at the end. Chinese ballistic missiles as hard to intercept as ICBMs (harder, actually, since they can dodge at tens or hundreds of gs), yet precise enough to hit a ship directly, can prevent American aircraft carriers from operating anywhere within a few thousand km of mainland China.

      1. To elaborate on the revolution this provides over ICBMs:

        There is a reason long-range ballistic missiles are tied so closely to nuclear and chemical weapons, and it’s not (just) the economics of such an expensive delivery system. Early ICBMs had error ranges on the order of a kilometer; modern systems, with mid-course corrections and a bit of aerodynamic maneuvering in the reentry vehicle, can get within maybe a hundred meters. Even non-fusion nuclear weapons are underpowered if you’re trying to damage a mildly-durable target with those near misses, and with the cost of nuclear weapons and long-range missiles that limits them to low numbers of high value hard targets or to the soft targets of deterrence..

        Hypersonic glide vehicles change all of that. Suddenly you can lob a missile a couple of thousand km away, and expect to hit a carrier or cruiser on the other end. You can use less costly (monetarily and politically) chemical explosives, throw them out at every little blip on your radars, and generally shift the balance of power towards the ground-based missile launcher away from the ship or airstrip.

        1. political cost : I tend to disagree, on the receiving side, how do you decide if this missile salvo that was just shot is ICBMs going for a decapitation strike or an anti ship missile aimed at a carrier ?

          You have less than 15 minutes to make a decision. More like 5-10 minutes if you are Russia / China (less good detection capacities – shorter decision loop).

          Anything that can conceivably mistaken for a nuclear tipped SLMB / ICBM is a non starter between nuclear armed powers.

          1. That will be an interesting problem as hypersonic are pushed. The decision time will be much less than 5 minutes. The weapons won’t have the slow startup of traditional ICBMs. Gun launched hypersonic weapons will be even faster to target

          2. The escalatory consequences of very short decision loops are not a trivial matter, but neither are they a new one – we’ve had ballistic missiles since the 1950s (or 1940s if you speak German).

            And, once again, hypersonic missiles are *slower* than ballistic missiles of comparable range. Not much slower, but somewhat so. The public reporting on this has been abysmal.

          3. Everyone is dumping large amounts of money into hypersonics.
            So anyone thinking they are just new fancy ICBMs is mistaken. Most of the information on the efforts are classified so that explains the lack of decent info in the press.

            They are extremely maneuverable and don’t have to leave the atmosphere like traditional ICBMs. Their ability to operate at low altitudes makes them much harder to defeat. Also they have a much smaller form factor than a traditional ICBM so hard to track. A gun launched hypersonic weapon will be hard to identify in time to respond.

            The railgun program that was just cancelled claimed to have enough control to hit a specific building at 210 NM. That was the claim 10 years ago so just imagine where it is now

          4. > The railgun program that was just cancelled claimed to have enough control to hit a specific building at 210 NM. That was the claim 10 years ago so just imagine where it is now

            I’d imagine that it is now at *less* than that. They figured 10 years ago that this level of performance could be possible, tried to do it, found out that feasible performance was much less, and that’s why they cancelled the program.

            I believe the abysmal reporting John Schilling mentioned is not a matter of reporters writing “we don’t know” because the accurate information is classified, but a matter of reporters writing false/misleading/nonsensical things because they are confused. For instance, the claim that hypersonics are “extremely maneuverable” is very commonly repeated but it’s at best misleading, since it’s usually phrased in a way that brings to mind the maneuverability of an aircraft. Indeed, most of the confusion is in thinking of hypersonics as if they were just like aircraft except faster, since they are not. It’s often more helpful to think of them as a bullet that can deploy (and cut) parachutes as they choose. At hypersonic speeds, the maximum possible lift-to-drag ratio (i.e. how efficient a wing can be) is terrible, thus a missile will lose a disproportionately large fraction of its speed in any turn. This is on top of the air resistance of flying in a “straight” (ballistic) path, which is already enormous and tends to impose quite a bit of deceleration on the missile on its own. Separately, even at silly accelerations (pulling dozens of g) the hypersonic speed works out to ridiculously wide turning curves. Hence the bullet-with-parachute metaphor: start on a (ballistic) path that would overshoot the target, then briefly deploy a parachute for just long enough that with the resulting *slower* speed, the new (also ballistic) path hits the target. Yes, this can include a bit of sideways course adjustment as well, but this is an adjustment, not the gross change that “maneuver” brings to mind. The capability is not to fly in a circle and hit the target from some direction they don’t expect to, but to send a missile roughly, but not exactly, on the ballistic path, and thus evade a defensive weapon that could throw a countermissile into the face of a missile that exactly followed the ballistic path.

          5. No. The issues with with the railgun were political / supply related and not projectile or launcher related. It was slated for DDG-1000 and when that order was drastically cut they had to look at other ships. The rest of the fleet didn’t have the power available (nuclear for steam not nuclear for electricity) so they had to rely on capacitor banks. That started taking up too much weight / space so the requirements kept getting pared backed till it wasn’t useful. At the same time hypersonics started taking off and looked like they could fulfill the mission. I am sure the railgun guidance technology transitioned to the hypersonics programs.

          6. I was referring to the political cost once the strike has been launched and hit (or missed) its target; nuclear weapons have a special escalatory quality.

            To your point, though, the confusion with an ICBM launch is indeed a problem; that’s a big reason why the US abandoned much-cherished plans for HGV-tipped missiles with intercontinental range (under the name “Prompt Global Strike”). However, it’s not a problem for of the type of intermediate-range system China is looking at for anti-ship purposes. Those are very distinguishable on early-warning systems from hotter and larger intercontinental launches; they look like a Pershing II or a DF-16, not a Minuteman II or a Topol.

        2. In addition to ballistically boosted glide vehicles, research continues on developing air-breathing missiles capable of high-Mach flight.

        3. I would not count on hypersonics hitting a carrier or other moving target any time soon. Moving at a couple of kilometres every second in atmosphere makes it very difficult to change course. There’s also a plasma build up around a hypersonic vehicle that blocks most radar and other guidance signals.

          The space shuttle was hypersonic on re-entry, and an astronaut (think it was Chris Hadfield) talked about how this changed flight planning. Like an airliner, the pilots had to plan for diverting to another landing field if necessary. Because the space shuttle was moving so fast, the alternative for the USA might be in Australia! They also couldn’t talk to ground stations until they’d slowed down to “only” being supersonic.

          Space shuttle Columbia also showed how vulnerable a hypersonic vehicle in air is to even slight damage 🙁 hence the interest in lasers.

          I’m not saying that hypersonic missiles won’t work or can’t work, just that it won’t be as easy as some proponents think. And this is just technical issues, as @folbec points out there are some serious political implications too.

      2. I’m no expert in this area, so maybe there’s an easy answer, but wouldn’t missiles precise enough and fast enough to deny the western Pacific to American carriers also be capable of denying the Straits of Taiwan to Chinese troop ships? Maybe the Chinese or the Russians have a temporary advantage in technological innovation in this area, but that surely won’t last.

        1. Which would be an argument for China to seize the initiative while they have a temporary advantage, no?

          1. Yes, if you are willing to take a big risk. I don’t know if that characterizes President Xi. Also, it depends on whether you view the long-term of correlation of forces as working in your favor or not. I don’t know Xi’s thoughts on that either.

            Maybe Prof. Devereux should give us his views on the “Thucydides trap”: are wars often caused by an established power that feels threatened by a rising power? Lately I’ve encountered an alternative reading, that wars are more often caused by a rising power that feels itself hemmed in and unable to rise further unless its adversary is neutralized. As applied to the Peloponnesian War, the first theory implies Spartan “war guilt,” which was Thucydides’ claim (not that he or any other ancient Greek would have used that phrase), the second implies Athenian responsibility.

    3. Hypersonic missile are kinda a stupid idea (cruise missiles traveling 5x the speed of sound). They (1) evade air defenses at rapid speeds, (2) are expensive, (3) fragile, and (4) we already have better technology already.

      (1) Any weapon that is fast enough to generate stratigic surprise must be treated as a nuclear strike
      (2) The amount of energy and technology to create a global strike cruise missile is insane.
      (3) Global strike through the atmosphere or near orbit is… highly vularable to changing conditions and exposed to detection and interception to some degree.
      (4) ICBMs travel at 15x to 30x the speed of sound, can hit any target on the planet, and re-enter at an angle that makes it functionally impossible for any known defense system to work against.

  5. Static defenses are alive and well in the form of minefields (both landmines and sea mines). In some ways they are TOO successful, being more expensive to remove than to plant. A variety of methods is used, including kickstarter project Mine Kafon (essentially a militarized tumbleweed / perekotypole) and trained rodents (“Magawa, a giant African pouched rat, was awarded the PDSA’s Gold Medal for his “lifesaving bravery and devotion” after discovering 39 landmines and 28 items of unexploded ordinance in the past seven years, according to the charity.”).

    There are mine trucks and mine launchers which can deploy a field within minutes. These may not be particularly well hidden (which is a plus from a humanitarian perspective) but with with current drone development it will be a short time before mines reorient themselves and drill in after impact.

    Also boobytraps and remotely detonated explosives.

    Perhaps the way Bret Devereaux framed this series (“Fortifications”, rather than “static defenses”) made him overlook the above.

    1. “Fortifications” is generally used for defensive structures that keep the human occupants alive, and are intended to have at least some kind of garrison if not a permanent population. A minefield can be part of the static defence for a fortification, but by itself is not used for such. Neither are remotely detonated explosives and other traps. So yes the series is framed in such a way as to exclude them.

      Minefields are also used as offensive systems, particularly at sea. The US conducted a minelaying campaign around Japan in late 1945, and for many decades kept the post war analysis secret because minelaying had been so destructive and so cheap that they didn’t want to give potential enemies ideas.

      As for self-digging mines, I doubt it. Ask any farmer: digging holes, planting things in the holes, and covering up again takes a lot of energy! Much better to concentrate the power source and mechanical digging tools in one big machine that plants mines.

    2. Per a military engineer friend of mine (and my own reading) – a (land-based) minefield, like any other obstacle, is considered useless if not covered by observation and fire. Now, they were speaking of a modern-system force in a peer conflict, but the underlying point is that if you are unable to prevent the enemy from sweeping the field/clearing the obstacle, it may as well not exist, because sweeping/clearing is cheaper than laying/emplacing

      1. A minefield would be limited to slowing movement down. Either it’s cleared via equipment, or there are some armies out there callous enough to use people to clear laneways..

        1. If you can’t take advantage of them slowing down, you have spent resources (mines and layers are not free) without costing your opponent equivalent resources. It’s not that difficult/expensive to sweep land-based minefields.

          1. “Warfare is the efficient use of time and space, lost space can be recovered, lost time? Never.”

          2. “without costing your opponent equivalent resources”

            Even an artillery-delivered minefield is not without cost to lay (in both time and logistics, and quickly cleared with bare-minimum engineering effort. Any serious minefield requires more effort to put in than remove.

            Not to mention the whole point of obstacles and barriers is to force the enemy to go around them while under fire. Without the “under fire” part, why lay them in the first place?)

            Outside of the Korean DMZ, and some very specific locations, you can’t actually deny land passage with minefields or obstacles, and even the DMZ is not impassible with the lavish use of minefields and obstacles (plus, they’re covered by fire and observation).

      2. Now I’m envisioning semi-robotic drone mines that can sense approaching forces, do IFF on them, radio an alarm if they detect hostiles, activate sentry drones armed with machine guns & rocket launchers, etc.

        1. “The dinochrome brigade is sending a Bolo. Hold what you got, boys, the cavalry is coming”

  6. Typos:

    “which fits with the modren system,” – modern

    “even as the close fight behinds.” – Can’t tell if this is some usage I don’t get or an error.

  7. Seeing the american fort in the last section makes me wonder if it’ll all come full circle with future armies being fuedal levies fielded by billionaires from fortress mansions.

    1. It’s also really easy to overstate the power wielded by private individuals. It is considerable, but almost insignificant when compared to the overwhelming might of the state.

      We sometimes misunderstand that because we have normalised and bought into the ideology of the nation-state.

  8. “One wonders how central these sorts of fortifications will be in the history of the American limes; will archaeologists marvel at the striking similarities in design between American FOBs located many hundreds of miles from each other?”

    Future historians will be less amazed, because their field will be very different. Rather than seeing these artifacts and having to infer the (likely-oral) Roman doctrine behind them, they will have access to the very field manuals you’re citing, written in trivially-copiable formats and extant in the millions.

  9. Interesting comments about the elastic defenses. Most of my exposure to military history comes from wargames, along with books suggested to me through the wargames I’ve picked up at some point or another. But all of those games are themselves products of thinking contemporaneous with the modern system, I think the oldest wargame I play semi-regularly is Bar-Lev, from 1977.

    That makes me wonder how much modern system thinking colors perceptions of older eras of warfare, because in all of my gaming, elastic defense is very much a thing and something you try to do unless and until you’re so battered down you can’t initiate any sort of counterplay at all. But is that a set of anachronisms? Has elastic defense been more active as you get closer to the modern systems of fighting? I’d be interested in some sort of mapping, although I realize that’s a lot of annoying work and I don’t want to dump a chore on anyone.

    Also, happy new years!

    1. Elastic defense depends on observation and communication. When you’re playing a wargame, you generally have complete observation of the whole field, and often instantaneous response to your orders from your units, so it’s possible to do even when you’re e.g. playing out the Roman legions lost in Germany.

      A somewhat more accurate way to play a wargame set in earlier times would be to have a team of commanders on each side, situated in separate rooms, where each player can only observe his own portion of the battle. A commander would have to write orders and have them sent between the various rooms, with the possibility of the orders being intercepted, lost, etc. The units would only move when the orders were received, and finding out whether they moved would require a communication back to the commander.

      Of course this would require a >lot< more players and would likely be less fun overall.

      1. A long while back, I actually played a “Matrix game” like what you’re describing, over a fictional skirmish in North Africa during the Sonnenblume. 6 players to a team, along with a bunch of referees for things like communications and determining results of the various player actions.

        My unit scored the highest of anyone’s in the game. I also got court-martialed for refusing a direct (stupid) order.

      2. I’ve both run and played “double-blind” tabletop wargames, though in those games each side has had perfect knowledge of what “their” units could observe and perfect C3 between and among “their” units. (These have all been “modern,” “near-future,” or “far future” worlds where networked units are not only plausible, but not having networked units would be implausible)
        Computer wargames have embraced the notion of Fog Of War, though most of the mass-market ones still project the “modern system” onto their rules set.

        (Note that the entire Tabletop Role-Playing-Game industry grew out of a desire to simulate imperfect knowledge and communications)

        1. “not having networked units would be implausible)”

          For hobbyists this is OK, but for military this sounds like a recipe for disaster if the networking fails for some reason.

          Yes, the networking will fail – from jamming, to enemy action, to something as simple as a bug. Note the bug currently jamming Microsoft Exchange Servers, caused by poor naming practices on malware scanning updates…

          1. “implausible” is a wonderful weasel-word, especially with military science. Implausible yet possible things will bite you in the ass.
            As Microsoft is learning, again; and I, as a humble sysadmin, have learned countless times

      3. There is actually a famous classic W40K batlte report (its old enough that they made destroyed tanks “smoke” using hamster bedding!) where they played something like that. The commanders could only talk to each other via a limited number of notes, and a single short sentence via recording device… That was also played up for thier opponents.

      4. A wargaming forum played a Kriegspiel-type game called “The Flight of the Eagle” based on the Napoleonic wars, with a referee tracking what each commander on each team could see and modeling things like how long it takes a messenger to travel between players. It took eighteen months to finish.

        The Kriegspiel That Would Never End:

    2. I recall from his posts about castles and various LOTR sieges that it is unrealistic how few sallies defenders in these things attempt, plus discussions of bastions and features that make things easy to defend but also easy to retake. So I’d say elastic defense has always been something of a thing.

    3. Elements of elastic defense can be found in ancient warfare- see Dr. Devereaux’s series on the Siege of Gondor, for instance, which discusses the use of deliberate counterattacks to disrupt a besieging force.

      Besiegers have been building lines of circumvallation around places they wish to assault for at least two thousand years, precisely because while the main line of defense for a walled city is necessarily the walls, there is nothing stopping the enemy from sallying out of the walls. This can be very inconvenient if, for example, you are trying to build a siege tower in preparation for an assault and the enemy sneaks out and sets it on fire in the middle of the night. That’s elastic defense right there, in a way.

      Likewise, the entire concept of using fortified points that act as bases for raiding (preferably cavalry) forces that can strike at an enemy’s supply lines or prevent them from dispersing to forage? Elastic defense, in a way.

      The trick is that, as Mark pointed out, elastic defense relies on observation and communication.

      Observation because it can elastic defense be carried out over in areas where the defenders have a clear idea of what is happening. After all, defenders are usually outnumbered, and attackers have a very strong incentive to lure defenders out of any fixed fortifications they may have. As such, the defender in an elastic defense scenario has to be extremely careful to avoid having their own forces drawn into a trap and ambushed outside the cover of the fortifications. This requires excellent situational awareness.

      Communication because you must be able to direct your own forces against the weak points the enemy exposes as they deploy and prepare for their own attack. So not only do you need to gather enough information to be reasonably sure you’re not sending your ‘elastic’ counterattackers into a trap, you need to issue them instructions to act on weaknesses that may be transitory or may not be visible to the troops on the ground. At the same time, you also need a loose and professional command structure, because your subordinates need to be able to exploit weaknesses you can’t see, or inform you of their existence.

      So pre-modern elastic defenders, lacking aerial reconnaissance, radios, and so on, normally have to employ such a strategy on a scale of “over an area no larger than you can see from standing on a tower in your own fortress” or with very small, often squad or platoon-level raiding forces.

      1. One example of sucsessful counterattack, albeit not by the besieged, is the cavalry raid on the Turkish camp during the site of Malta in 1565.
        The horses of the Knigths of St. John were transferred to Mdina before the siege and the attack of the cavalry upon the rearguard on August the 7th break havoc in the Turkish army, so that the attack on Birgu was cancelled just when the Janissaries have driven back the defenders from the walls.

      2. Or you have to trust your subordinates to be able to act on their own, without communications from you.

  10. Thanks! Lovely piece. Two quick notes:

    1. The German infiltration tactics of WWI turned out to be a tactical dead end, once the defenders realized the proper countermeasures, which were for the fortified hard points in the trench system to hold tight and defend their spots, the artillery to hammer No Man’s Land to prevent German reinforcements, and counterattacks to sweep away the lightly armed and limited number of storm troopers (similar tactics worked pretty well against armored assaults in WWII). The storm troopers got bogged down and, without reinforcements, could be swept up pretty effectively.

    The British and French tactical approach (a range of things, but mostly encapsulated under the phrase ‘bite and hold.’) by contrast never got solved by the Germans. The central premise was that if the Germans were going to give the front line away and then counterattack, why not take the front line, fortify it, and then slaughter the counter attackers? That’s what the Entente forces did during the 100 Days — launch a heavily armed assault to take the German front line, set up quick defenses there and range the artillery barrage into the gap between 1st & 2nd lines. Once the counterattack had been slaughtered, and supplies built up, repeat the process. The odd part of this was that the British & French first wave infantry were actually much more heavily loaded down than previously with what they needed to set up those defenses (machine guns, mortars, barbed wire, etc).

    The problem for the Germans was that the obvious response was to build up forces in the front line to prevent the first bite, but then the British and French artillery would hammer them directly. So they were stuck, and the Entente pushed forward steadily in August-November 1918.

    2. The modern American FOB relies on the fact that the enemy doesn’t have (and couldn’t really use) the kind of heavy artillery that made such fortifications problematic. A Taliban ‘Big Bertha’ would have been quite destructive.*

    *and pretty quickly destroyed by counterbattery/air strikes of course

    1. German infiltration tactics weren’t a dead end simply because some of their core concepts, particularly Auftragstaktik, remains core to the infantry doctrines of the major English-speaking militaries and it isn’t hard to find the seeds of later German maneuver war (bewegungskrieg) in their late-WWI infiltration tactics and the indebtedness of AirLand Battle to bewegungskrieg is pretty clear.

      You can argue they’re ineffective, but you can’t say they were a dead end.

      1. It might be better to say that they were very much not a dead end as far as “evolution of modern infantry tactics” was concerned, but very much were a dead end as far as “figure out a way to reliably pierce trench networks when both sides have 1917-18 technology to work with” was concerned.

        Of course, as you pointed out in an earlier series, Dr. Devereaux, pretty much everything anyone tried was a dead end by that standard. And as noted above, the closest anyone came up with to a ‘solution’ within the confines of the era was to just give up trying to break through and find a suitably efficient ‘bite and hold’ strategy that could seize territory at a an acceptable ratio of attrition. So long as, that is, the attacker was content to expend artillery ammunition and manpower just to advance their own trenches several hundred meters at a time with each ‘bite’ and then inflict additional casualties on the enemy by ‘holding’ against any counterattacks.

        1. “It might be better to say that they were very much not a dead end as far as “evolution of modern infantry tactics” was concerned, but very much were a dead end as far as “figure out a way to reliably pierce trench networks when both sides have 1917-18 technology to work with” was concerned.”

          Exactly, which was the point I was making (so I probably shouldn’t have made a remark about armored warfare as an aside, though even there, bewegungskrieg collapsed pretty quickly against an enemy that knew what to do, defensively).

          “by that standard”

          And ‘bite and hold’ did, in fact, lead to breakthrough eventually because the Germans couldn’t build trench systems fast enough to keep up (or reconfigure the railway lines that supplied the trench systems).

          But remember that the goal is not to reliably pierce trench networks but to win the war, and the bite and hold tactics helped do that quite effectively.

          (Militaries and military historians have tended to fall in love with German methods and neglect the methods that in both world wars that actually *won* the war, for all sorts of reasons).

          1. Not this again..

            The strategic methods that “won the war” for the Allies are the ones that historians are in love with–mass production, industrialization, not being tremendous jerks to the occupied populace, actual strategic coordination, etc.

            Tactical and operational methods can either help or hinder your strategy, and the fact is that German tactical and operational methods made fundamentally unworkable strategies seem like they could work, which indicates that they were pretty good. (It also helps that Blitzkrieg is flashy, while time-on-target artillery barrages are less so.) The Allies were almost certainly the Germans’ equals at both by the end of the war, but they spent until late 1943/early 1944 playing catch-up.

          2. If memory serves me right, the French developed bite and hold tactics as early as 1915? The Germans very quickly developed an effective response, which was to train their troops in retaking their own trenches. They would hold mock counterattacks on their own front lines, so that when the real attack came, they were well prepared to quickly dislodge the attackers before they could fortify the recently captured positions, and move their artillery forward.

            I suspect that the reason bite-and-hold tactics were successful during the 100 days campaign, is that the effects of attrition prevented the Germans from being able to mount proper counterattacks. Although it’s possible other conditions/developments were at play too?

          3. I’ve said it before, and I’ll say it again. From at least WWII on, the American Way Of War has been to “use land forces to locate and pin down the enemy for long enough to beat them to death with Our Awesome And Terrific Logistical Power.” This is an expansion of the American section of the old saw about “how to determine the nationality of an unknown unit in WWII by seeing how they respond to being fired on” which describes the American Response as “nothing happens for a few minutes, then your position is obliterated by artillery/airstrike.”

            Time on target is not sexy. It’s nerdy as all get-out. And it works

          4. @60guilders

            Yes, that’s the narrative that I was referring to: the ‘Germans were great at tactical & operational stuff and the Allies only overwhelmed them with mass.’ The American Civil War equivalent is of Lee being a tactical genius overcome by the mass of Union soldiers. Both have some truth to them, but are terribly overstated and distort the history.


            What the French (and British) were doing in 1915-16 wasn’t really ‘bite and hold’ as the 1918 forces tried. They were still trying to break out of the trench system, rather than immediately hunkering down and fortifying the front line they had just occupied. That’s why the 1915-16 attacks were vulnerable to German counterattacks and the 1918 versions actually welcomed them.

            Anyway, thanks for the discussion (and to Bret for the post). I’ll leave it here.

          5. @silbey

            I guess it depends upon what your definition of bite and hold tactics is? Because I happen to have an old Strategy and Tactics magazine at hand, while it doesn’t use the term bite and hold it does say:

            “The French did discover that if they were content to simply take the first line of trenches, they could cut their losses considerably. Even this could be nullified by the German practice of holding trained reserves, which had actually rehearsed the recapturing of their own first lines. It was this counterattack technique that made the massive assaults between 1915 and 1917 completely useless.”

            According to the wikipedia page for the Battle of Amiens(1918) (part of the 100 days offensive), after the initial success on day one, the advance quickly slowed, partly because they had outrun their artillery. They had gained something like 13km (in some sectors) on the first day. So if they weren’t digging in, after taking the first line, and waiting for their artillery to come up, is that still bite and hold tactics?

            I didn’t really think the 100 days offensives were based on bite and hold, but maybe people use different definitions?

            Thanks for the conversation.

          6. @silbey
            I don’t disagree (that was why I included the bit about the Allies improving their operational and tactical methods), but I grow somewhat vexed with the “correct the historical record vis-a-vis relative performance not by figuring out that the side that is denigrated was better than we thought but by deciding that the side that is praised was actually terrible at what they were supposedly good at.”

    2. The 100-days offensinve worked, because of fresh supply and soldiers from the US. It’s really hard to overemphazies the effect of the war entry of the US.
      Without those 250k fresh frontline troops every week, let alone labour battalions, cheap coal for the railroad, ammunition etc. pp. the 100-day offenive would have ended, like every offensive in that war.
      The infiltration tactics on the other hand, failed as our host wrote somewhere else, because no infantry offensive can move so quickly, that it can overtake the railways deploying reinforcements for the counteroffensive.

      1. That depends on whether the reinforcements are actually available (a question both of numbers and morale). Also ‘bite and hold’ did not just over-run the first line. A quick couple of heaves took out the gun line (as at the storming of the Hindenberg Line), and artillery guns and ammo stockpiles are not quickly replaced. The Germans lost over 6,000 artillery pieces in the 100 Days Offensive.

        Second counter is sequential offensives (also employed to great effect by the Red Army after 43). Launch attack, take bite, hold, draw in enemy reserves, launch second and then third attacks far enough away that reserves need to be entrained to counter. With good timing enemy reserves will be caught mid-travel, too late to respond to either point (as happened, for instance, to SS formations just after Kursk – sent south to meet an offensive on the MIus, then north to try to hold Kharkov, before going to Italy). In the Hundred Days the Allies did just this.

        1. My point was, that the 100 Days offensive could do what it did, because the US gave the other Entente Powers men and material to outspend the Middle Powers.

          The whole thing worked because the Entente was able to open up new fronts in the Balkan, and in the middle east, stabelize Italy etc. pp. Which again forced Germany to divert troops to all of those theaters. That had nothing to do with tactics, and all with strategy and logistics. And now, ask yourself, why did they do it in 1917/18 and not years earlier? And the simple answer is, that access to American War production (and American Soldiers) allowed them to diverted ressources from the Western Front.

          1. The Entente powers were already outspending the germans by a significant marigin before the US entered the war (US was definitely important in allowing this to happen, but it did so without actually entering the war) especially in materiel it was a pretty lopsided affair

    3. Your point 1 was (independently? I have no idea, because I was not well-read at the time) evolving in a RTS game that I was playing in the nineties ( Total Annihilation), where short-range “hardpoints” backed by long-range artillery and supported by a mobile reserve were “the meta” for building “unengaged” defenses.

      (TA was weird in that it allowed for some seriously long-range artillery compared to other games in the genre at the time, and had a very conceptually different economy than, EG, Starcraft, in that the resources “streamed” and were unlimited in-game)

  11. “Torpedo shells”

    Ah, the wonders of semantic drift. When Farragut damned them (“full speed ahead!”), they were mines.

  12. My understanding was that at least as recently as the GWB administration, anti-missile tech still didn’t actually work—evidence to the contrary being mostly staged dog-and-pony stuff, like they’d actually put a radio transmitter on the target and that was what was actually being tracked. (I’ve not heard any subsequent reporting about it.)

    Have there been advances in the subsequent decade that make these finally feasible?

    1. Yes, to put it mildly.

      More generally, anti-ballistic-missile tech (henceforth ‘ABM’) has actually worked for quite a long time… as long as you were willing to put a small nuclear bomb in the anti-missile. This gives you a big enough blast radius that you can accept a “meh, close enough” solution. Get your countermissile into the general vicinity of the incoming nuclear warhead for even the tiniest fraction of a second and BOOM, it’s gone. You’re still slinging around nukes, but at least you have a two-kiloton explosion happening high up in the stratosphere, which leaves you comfortably out of the blast radius, instead of a two-megaton explosion happening in your living room, which doesn’t.

      Development of ABM systems that work without a nuclear warhead is rather more challenging, because you need to get the interceptor missile much closer to its target, and the timing is much more precise because you have only a very narrow window of opportunity if you want to score a hit and take out the incoming warhead. Making this viable means a lot of development, prototyping, and intermediate steps to go from napkin-math design concept to a working weapon system.

      This runs into the issue that not all tests of a new weapon system can, or should, be under ‘realistic’ conditions. For instance, suppose you are trying to design a heat-seeking missile that can be shot out of a cannon and home in on the enemy. I don’t know why you’d want it, but suppose you are. There are several key component requirements: “heat-seeking,” “can be shot out of a cannon (without breaking),” “home in,” and “on the enemy.” Importantly, each component must at some point be tested separately.

      At some point, you will probably perform a test where you try to launch the missile out of something more gentle and delicate than a cannon, as a crude test of certain functions. You will probably probably test the heat-seeking warhead on targets that are far easier to detect than a typical enemy tank, in order to differentiate “the heat-seeker broke during launch” from “the heat-seeker functioned after launch, but was insufficiently sensitive to detect the target.” There is no way to tell the two scenarios apart reliably, after all, aside from making intentionally easy-to-detect targets and firing at them.

      All of this is part of the testing program for this weapon system, and all the tests must be performed at some time during development except in the very unlikely event that the very first prototype version of the weapon performs perfectly and exactly as expected in all ways.

      But God help you if someone reading about your test program goes “hey, wait a minute, you test-fired this heat-seeking missile at a cardboard cutout of a tank that had a bonfire set behind it? That doesn’t sound like a very realistic field test, har har har!” They may very well be saying this about you decades later, long after your heat-seeking missile has successfully been used on the battlefield and shown to work just fine, thankyouverymuch. 😛

      I wouldn’t be surprised if the GBI ABM system was tested against targets with a radio beacon built in. I can think of several reasons to do this. For example, to confirm that all the parts of the missile besides the target-seeking apparatus work. Such as telemetry systems, such as the actual maneuvering controls that physically steer the missile, such as any warhead that must detonate at exactly the right millisecond to destroy the target. All these systems can best be tested under conditions where it is a ‘controlled variable’ that yes, this particular test missile will in fact get acceptably close to the target and will not lose track of the target.

      Which means doing a batch of dummy tests against targets with a radio beacon built in.

      But the fact that the GBI system continued its development and was deployed in the field reflects the fact that this testing phase (during the Bush administration, well over a decade ago) is now over, and the system is deemed ready for prime time. Or at least good enough that having it deployed to a large military base in Alaska is worth the cost of maintaining a base in Alaska in hopes that it’ll shoot down a few missiles if North Korea gets mad at us or something.

      Field deployment is too expensive to be done with a weapon system that you know doesn’t work and needs refinement.

      1. You underestimate the willingness of the US government to believe its own propaganda, especially in the post-Bush II era. Us physicists keep explaining why non-nuclear missile defense won’t work, and the military keeps doing strawman tests under unrealistic conditions and saying “look, yes it does”.

        1. I’m sorry, but “you physicists” are wrong, and have been for a decade plus, unless there is a top secret group in the MDA pulling off a “moon landing hoax” level conspiracy theory on the thousands of military personnel and contractors involved in these “strawman” tests.

          GMD has a mixed record but has still intercepted realistic tests of realistic targets. Aegis ABM has been extremely successful, with about 3 dozen intercepts with a hit rate of around 80%. The more recent of these tests have been performed with tactical weapons, tactical sensors, and tactical crew. The only “unrealistic” part is that obviously the shooters know they are participating in an exercise (but the exact time and details of the incoming target are unknown to them). THAAD and PAC-3 have also successfully scored hit-to-kill intercepts of ballistic missile targets at a high success rate under similar conditions.

          1. An Aegis 80% hit rate is great and all but in a general nuclear exchange oh a 100% hit rate is acceptable. An 80% hit rate against 10,000 Russian war heads is the end of all life in North America.

            Which is a fundamental problem for nuclear missile defense. Anything less then 100% perfect is functionally useless.

          2. “Useless … in a general nuclear exchange with the (ex)soviet missile fleet” is FAR from “not useful.”
            There’s a saying about the perfect being the enemy of the good enough that applies here. Among other things, the chinese missile fleet capable of hitting the US is (at the most generous, yet plausible, estimate that I could find) somewhere around 100 missiles and 200-300 warheads.
            Their shorter-ranged fleet is much larger, but also easier and cheaper to defend against.

            (Also, depending on where the Chinese missiles are based and what our local diplomatic situation is, we might be able to station Aegis systems to target Chinese ICBMs in their boost phase, where it’s a much less complex problem to intercept them)

            Without knowing more about that “80% success rate” and how it’s derived and what it’s measured against, I won’t even go so far as to say that Aegis is worthless against a general nuclear exchange; it all depends on how many Aegis installations we can build and how many countermissiles we can launch.

          3. Not to mention small-fry nuclear nations like North Korea and Pakistan.
            (Not that I expect Pakistan to nuke the US, ever, and I doubt they even care to develop ICBMs when very short range projectiles can hit every one of their strategic enemy targets from Kolkata to Rajkot and Madurai to New Delhi.)

          4. If the Chinese have 200 war heads that can reach the US mainland, which given the current investment by the CCP in new missle silos isn’t a given ( ), a 95% kill rate of inbound war heads is 10 thermonuclear strikes on the United States.

            No matter how good missle defense gets, it’s still functionally worthless. 10 thermonuclear strikes doesn’t end human life in North America, but it would probably be enough to end the political entity known as the United States.

            And, with 5% effectiveness being 10 thermonuclear strikes, at 90% effectiveness you’re talking about 30-60 million people before starvation and social disruption.

            In most things the perfect is the enemy of the good. In nuclear warfare, perfect is the only acceptable outcome. Which is why any political or military decision that leads any semi-rational decision maker to believe that a nuclear war can be won or risked is a horrifically awfull path.

        2. For the past seven years, the government of Saudi Arabia has been participating in an actual war that has resulted in I think about two hundred ballistic missiles being launched against targets in Saudi Arabia and/or parts of Yemen under Saudi protection. These aren’t “strawman tests under unrealistic conditions”, and yet the Saudi armed forces keep shooting down the missiles.

          Last time I checked, they were getting about 75% effectiveness at shooting down ballistic missiles targeted at areas within the engagement envelope of their missile defense batteries. That’s not the optimistic 90% claimed during e.g. Desert Storm, but neither is it “missile defense won’t work”.

          Extrapolation to large-scale nuclear war between the US and China or Russia is difficult for a number of reasons, not the least of which is that the current US missile defense systems aren’t designed for that sort of war. But it’s not the no-brainer obvious “can’t work” that is being claimed by “us physicists”. And it’s also not a physics problem but an engineering one, so physicists as such have no particular credibility in claiming they have the ultimate answers.

          1. True. But important clarification: many of the missiles the Saudis intercepted were cruise missiles, and none of them were ICBMs. ICBMs are particularly difficult to intercept – their boost phase is going to happen very far away where you can’t intercept it, their midcourse phase happens very, very high up where it is very hard to intercept and their reentry phase is extremely short and extremely fast.

            This makes ICBMs a much tougher target than the sort of missiles or rockets in use against Saudi Arabia or, say, Israel.

          2. I do not ethat I dont think the ABM systems were ever even theoretically supposed to defend against a russian total nuclear war scenario.

            They were however designed to maybe be able to stop a north korean nuke.

          3. True. But the missile attacks are small salvos – comparable to what Hamas occasionally launches from Gaza. A true mass attack (say – what Iran would launch, or Hezbollah on Israel) would overwhelm the defence. Which is why Israel avoids annoying Hezbollah and the Saudi don’t poke Iran too hard.

      2. “suppose you are trying to design a heat-seeking missile that can be shot out of a cannon and home in on the enemy. I don’t know why you’d want it, but suppose you are. ”

        I could see using a low-powered charge to launch a missile that then lights off a ways away from the launcher in an attempt to spoof where the launcher is. Not sure how well it would work – you would have to have enough energy to move the missile far enough away to be worth doing, while not having a large enough signature to render it pointless – but I don’t think it’s inherently unworkable.
        Interestingly you might then fire the missile in the “wrong” direction as part of the spoofing efforts…

  13. “Ironically, the collapse of the Liège forts convinced French high command that these fixed fortifications were unlikely to be held against determined German assault, leading to the progressive downgrading of the Verdun fortifications (their artillery was mostly removed for use amidst the trenches), despite the fact that the modernization efforts on the Verdun forts actually did render them relatively more resilient against artillery barrage, something the French would themselves find out when trying to retake forts Douaumont and Vaux after both were captured by the Germans in the opening of the Battle of Verdun.”

    This is one of those decisions that with the benefit of hindsight sounds stupid and shortsighted, but you have to wonder how the counterfactual would have played out. Would the threat of these fully-defended fortresses have stalled the German offensive? Or would they have been enveloped and blockaded as the Germans penetrated deeply into the weakened French field fortifications?

  14. Lots of responses by the time I finished reading, but some of these might still apply:

    Bret, first please look at this sentence beginning your second paragraph. I *think* that either “first of” or “was” will have to be deleted/?

    “look at the impact first of industrialized firepower…has on the design”

    Additional typos noted:
    came to were ‘fortress > was ‘fortress
    the same was as the main > same way
    each fort featured both smaller artillery > [both…and what?]
    where was substantial > there was
    city started in September, 1914 failed > [missing first coma to set off phrase and second comma ought to be placed after year/?]
    considerations the impact of tactical > [doesn’t seem to follow logically/grammatically/?]
    enemy access to that the army > so that

  15. “Artillery shells, alas, are not regularly grown by farmers out in the countryside.”

    Farmers in some regions of France would beg to differ. 😀

    1. “Liked” in the sense of “appreciated the mention.” The aftereffects of WWI (which were NOT repeated in WWII, at least not at that scale) are terrific and awesome in the literal sense. There are still places that are entirely forbidden (for reasonable safety reasons) for humans to go due to UXO both of conventional explosives and chemical weapons and will be for a very long time, measured in hundreds of years from now
      (The “regular” Iron Harvest is a “mild” version of this)

  16. The push for hypersonic weapons and long range artillery (i.e. SLRC) are going to cause issues with A2/AD. Hypersonic weapons are supposed to be too fast to stop. SLRC shoots from so far away there is no way to do counter battery.

    There is also an interest at present in operating inside of the enemy’s lines. Wargaming on how to survive and fight inside the 9 Dash Line has been common lately.

    Advances in 3D printing will also change what is needed in terms of supply lines. 3D printed artillery propellant will soon be a reality.

    1. I can see 3D printing being useful for printing parts for field repairs so you can reduce parts inventory, but I don’t see the advantage for printing artillery propellant although it is interesting that it is possible. Presumably 3D printed propellant is not chemically more efficient than regular propellant so there still needs to be a large mass propellant material to be sent through the supply line and then there is the need to actually print the propellant. And what about the rate of printing?

      1. 3D printing is about grain design. You can print web geometries that you can produce via other means. Control of the geometry allows for better control of burning and thus better performance. We are still a ways out before we test full size 155mm charges. That is about all I can say here.

        1. So this would be something like shipping a solid block of stuff (so perhaps less
          explosive and/or cheaper to make) that then gets broken down into the correct grain size etc. by the printer?
          Cool idea, but the block still has to be made and shipped.
          On the other hand, making the chemicals involved from the ground up seems like it would involve large amounts of energy since that has to come from somewhere.

          1. The block or spool or liquid (the exact method is part of the research) container is cheaper to manufacture and make durable than the final propellant grain. Depending upon the formulation some of it could be locally sourced. It is cheaper and easier to ship the raw materials than the finished part. The advantage for printing the propellant is that you can get better performance than traditional manufacturing. Printing it in the field is just an extra bonus.

  17. Like in so many other ways, the American Civil War was a “proto-industrial” war in that we see something like the progenitors of ring forts. According to a National Park Service website “By 1865, the Defenses of Washington included 68 forts, supported by 93 detached batteries for field guns, 20 miles of rifle pits, and covered ways, wooden blockhouses at three key points, 32 miles of military roads, several stockaded bridgeheads, and four picket stations. Along the circumference of the 37-mile circle of fortifications were emplacements for a total of 1501 field and siege guns of which 807 guns and 98 mortars were in place.”

    1. And much like the Belgiums would find, the moment the Union had a real general in Grant they pulled the garrisons from Washington because the troops were much more useful in the field then behind the fortresses. In 1863 Washington was the best defended city in the world, and by 1864 almost the entire garrison was in the field.

      1. It helped that by 1864 the idea of the Confederacy launching an attack on Washington DC was considered utterly ludicrous, after Vicksburg, Gettysburg, and Chattanooga.

        Also worth noting is that when Jubal Early launched himself out of the Shenandoah Valley in mid-1864, and attacked DC, an entire corps was pulled from the Petersburg siege lines and rushed to the city to reinforce the already-present garrison.

        1. Which just went to show how innefective a use of resources the forts were. Early was able to tie up a corp. but with rail roads and steam ships, the marching speed of an army, even one with internal supply lines, was already beaten by the mid 19th century mobility improvements.

          You really can see like 80-90% of the outline of the WWI trench stalemate in the American Civil War once you have equally sophisticated generals facing each other.

          Aside from artillery, which if the other technologies exisited it probably would have increased in capability very fast, the lack of repeating rifles and machine guns – thus requiring still close order formations – really prevented anyone from trying to intentionally cause a WWI style stalemate. But, you can squint and see the tech coming together.

  18. This might have been a question more suited to the previous article, but how did Napoleonic warfare circumvent fortresses so easily? I’ve noticed that sieges play a very minor role in Napoleon’s campaigns and he was able to take Vienna, Madrid, Berlin, and Moscow without much resistance by their fortifications, but he seems to have been operating before the major advancements you mentioned, such as explosive shells and useful indirect fire artillery.

    1. Hmm, Acre, Torres Vedres, Badajoz, Cuidad Rodrigo, Saragossa. But in general you are right. One answer is that the French Revolutionary and Napoleonic Wars involved much larger numbers, so fortresses could be masked and the main body go on. A second is that they manoeuvred by corps, along multiple routes (and maps and roads had both improved a lot in the late C18). A third is that the French relied on foraging more, where they could – so the ability of fortresses to interdict supply mattered less.

  19. Artillery shells, alas, are not regularly grown by farmers out in the countryside.

    Note to self for a fantasy setting: How would warfare in a world where, say, some kind of commonly-grown gourd was used as magical artillery shells be different? It would obviously depend on the nutritional value of the explosive gourds—if they’re a niche military good, or a decent supplement to a diet of bread and porridge, or if they’re nutritious enough to serve as a staple crop. Perhaps they’re supercharged with magic or something, which both makes them ideal staple crops and effective weapons, depending on how the local witch/military alchemist utilizes them?
    Worldbuilding exercise for another time, I think.

    But of course that statement was a damning one: a fortress which is only useful against enemies with substandard artillery and crippling shell shortages isn’t much of a fortress at all.

    Unless you’re designing it to defend against raiders who lack access to modern artillery and consistent shell supplies, but just reading that should make it clear how ridiculous that idea is.

    Playing the Clausewitz drinking game with US Army field manuals would be, by the way, a pretty good way to get really hammered really fast, but you’d need to know your Clausewitz, because the borrowings are typically unmarked.

    Sounds like a good ploy to get soldiers to study more military theory. Especially if everyone except the soldier who caught the Clausewitz had to drink!

    1. Ah, fantasy combat! Where you may have to deal with intelligent, self-propelled flame throwers that can also conjure food.

    2. I’m pretty sure that if a commonly grown crop could be used as a hand grenade with minimal chemical/magical treatment, all forms of civil society on a scale larger than villages would be quite unstable, because it would be too cheap and easy to equip yourself with weapons capable of blowing up the neighbors reliably. Or with blowing up the king’s tax collector.

      1. Fragmentation apples against the grain? If the weapon-crop literally grows on trees, conquest can be stabilized by cutting down the relevant trees, or otherwise messing with their production.

        Agriculture is an excellent source of plausible deniable magic. By this I mean that many ways to affect productivity to a significant degree (terra preta, or messing around with symbiotic or harmful insects, fungi, worms, soil microbes, etc.) are themselves outright invisible, or act in ways that are very unintuitive. Thus if a society should find (if only by authorial fiat) a practice that actually works IRL (see they will necessarily make up wrong, presumably magical, explanations (and generally focus on orthopraxy, as ancient religions did). If your village is almost but not quite at war with the neighboring village, you can sow ergot among their cereals under the cover of night, or perhaps cause their trees to wilt with a hammer, a nail and some water with fungal spores of the right kind.

      2. You might as well say that guns being common would make civil society on a scale larger than villages unstable, because then you could easily shoot your neighbors or the tax collector. You don’t blow up your neighbors because you rely on them in hard years, and you don’t blow up the tax collector because the king has way more melonpower at his disposal.

        1. You get out of civil society what you bring into it. There are nations where “society on a scale larger than villages is unstable” even though the kings have more firepower, because they don’t have legitimacy.
          On a more practical note – you can’t eat grenade pineapples. Even if they’re not outlawed, the subsistence farmer or sharecropper can’t afford to sow them, not if they want to eat through the winter.

    3. “…if they’re a niche military good, or a decent supplement to a diet of bread and porridge, or if they’re nutritious enough to serve as a staple crop.”

      Something something mustard gas.

      Happy new year!

  20. The “defense in depth” link leads to the first entry in the Siege of Gondor series, but it really should lead to the second entry.

    Unrelated: Edward Luttwak distinguishes among four types of defense: forward defense, preclusive defense, defense in depth, and elastic defense, where the last means the use of large, centrally-located armies which eschew fortifications (beyond the night camp) in favor of meeting and fighting any invader in the field.

  21. Going back to medival fortifications, since nobles pledged a limited time of military service per year to their superiors, has a siege ever been beaten by running out the service clock?

    “Sorry my liege, but our service time is almost up. If we can’t take the castle by the end of next week, we have to return to managing our estates.”

    1. I have no specific examples, but this was a pretty big problem for military leaders of the time. Imagine you are the German King waging war in Italy, were it often takes more than a month for your troops to even get to the theater.
      But there were ways around it. Yes, your knights only were required to deliver 40 days of service a year, but you could pay them to stay in the field. Well you could not SAY that you pay them, to stay in the field, those guys were nobles and being payed a wage was below their status. But you could offer them lands, or “gifts”. Many might have just stayed because of the chance of plunder.

      1. The 40 days was notional. That was unpaid service. Any German vassal signing up for Italy knew they were going for at least six months, more likely a year. Also – the whole ‘nobles don’t take money’ was much later (18th century). Medieval and early modern nobles were quite happy to take cash or even cheques (eg, the English exchequer valued the horses knight brought, and paid for loss on campaign. They also had set rates for knights and other ranks). There was a thing called a ‘money-fief’, where homage brought a salary instead of land. Quite common.

    2. I think legitimacy (which has also been talked about a bunch on this blog) plays a big role here too: If you think your king is a good king, has the right to rule from God, and generally his wars turn out good, you might be inclined to stay longer as a service to him, while if you thought he sucked and really his brother would be better, you might leave as soon as allowed/socially acceptable.

      1. Once succession stabilised in the major hereditary monarchies (England, France, Burgundy, Hungary), switching to a rival becomes quite rare, except where the succession fails. You could ignore or safely disobey a poor king, but replacing him was pretty hard.

  22. In a paper magazine, I have a recent interview with the professor David Frye. It’s about strategic walls, such as those built at the borders of USA and Mexico, at the borders of Belarus, the Hadrian Wall, Great Wall of China, the Maginot Line, Tres Long Mur – the wall in Syrian desert. He doesn’t distinguish between walled cities and strategic walls – he says they are very similar in practice. Apparently he’s promoting a new book (“Walls: A History of Civilization in Blood and Brick”).

    Some of the more interesting claims:

    “Plato praised Spartans, who banned walls, fearing they would grow soft. Aristotle thought this thinking was obsolete, because living behind walls is more comfortable and gives a better chance of progress.”*

    Voltaire supposedly called walls a monument of fright.(I’m unable to locate the quote – it was reported speech and not in English.)

    “It’s hard to find advanced civilizations and cultures without walls.”

    “The Manchu dynasty (China) was fond of repeating the Confucian wisdom: Trust virtues, not walls.”

    The Silk Road worked because it was protected by a wall.

    The North America was inhabited by nomad warrior tribes, and they mostly disappeared. South America was a civilization of wall builders, including the Great Peru Wall.

    Wall builders accumulated knowledge, produced writing systems, art. Peoples without walls produced military cultures lacking those.

    Walls have a stabilizing effect, because they prevent wars (Berlin Wall was used as an example, Kennedy – “A hell of a lot better than a war”). If someone’s looking forward to a war, he doesn’t build a wall.

    All great walls required despotic rules due to the immense amount of work required. Political/policing walls have a lower cost and private companies are very happy to build them.

    Up to Hadrian II, Rome based its security on client states surrounding it. Rome used a combination of corruption and military threats. Walls and fortifications appeared when it stopped working.

    (Great) walls work well at first, but then comes a false sense of security, and complacency. Doesn’t it sound like a version of the meme “Hard times create strong men.(…)”? Later, people often don’t care about defending them, or the walls fall by treason.


    *I find it strange that prof. Devereaux has written separate series about Sparta and fortifications and failed to mention their attitude. What prof. Devereaux does say about Spartan fortifications is very broad:

    “Spartan planning is both direct and unrealistic: find a choke-point, fortify it and hold it indefinately with a hoplite army.”

    “Put very briefly: Greek armies seem to have had relatively little carrying or logistics capacity. They did not seem to have generally moved with sufficient engineering tools or materials for effective field fortification or siege warfare. This is compounded by their inability to mill grain on the move (something Macedonian and Roman armies could do),”

    1. Now I’m no professional historian, but some of these claims seem a bit suspicious to me:

      > “It’s hard to find advanced civilizations and cultures without walls.”
      Sure, but not because walls are some product of “advanced” civilization, they’re a product of (and solution to) the common problem of “How do I stop my neighbor from attacking me?” Cultures all around the world (“advanced” or “primitive”) throughout history have converged on walls as the optimal solution to this problem given available constraints. Physically making it difficult for people to reach some place you want to keep them out of is a concept a 5-year-old could come up with, and Sone Age peoples have carried out again and again.

      > The Silk Road worked because it was protected by a wall.
      Uh…maybe partially, but certainly not for it’s entire length? Eyeballing it on a map it’s at least five times longer than the Great Wall of China, so even if the claim is that that was protecting it from nomads to the north that’d be at best maybe 20% true. I suspect the Silk Road “worked” because everyone along its routes had a vested interest in it continuing to work due to the economic prosperity it brought, and thus had reason to defend it. And even if the final link with China were to collapse it’s not like the intermediate stops had nothing to offer traders to the west.

      > The North America was inhabited by nomad warrior tribes, and they mostly disappeared. South America was a civilization of wall builders, including the Great Peru Wall.
      “Disappeared” how? Because there are hundreds of Native American tribes that still exist as political entities in the US and Canada at least, whereas the Inca civilization—as a political entity—is no more (plus, South America was hardly monolithic nor a single “civilization” at pretty much any point in its history). Also, there were plenty of non-nomadic wall-building polities in North America (the Aztec empire, for one—which is also no more—but also the Haudenosaunee), and nomadic ones in South America.

      > Wall builders accumulated knowledge, produced writing systems, art. Peoples without walls produced military cultures lacking those.
      Uh, ALL cultures produce art. (As Dr. Devereaux has carefully shown in his series on nomadic peoples—I hadn’t seen the kinds of clothes they made historically and they were pretty impressive.) Not all “peoples without walls” are militaristic: the Moriori people of the Chatham Islands were famously pacifistic, so much so that the Māori people from nearby New Zealand nearly genocided them in the 1830s without them fighting back, as they considered it a moral imperative. (And plenty of wall-building cultures have also been quite highly militaristic throughout time.) All (surviving) cultures also must accumulate knowledge (or they wouldn’t survive long in their environment!), passing it on through oral traditions in the absence of a writing system. Writing systems are certainly helpful for many purposes, but hardly necessary for a culture to survive, so it’s hardly surprising that not all cultures adopted one historically (the Inca empire [wall builders!] had at most a proto-writing system in the quipu, and they were doing just fine until outside disruption by the Spanish). I can’t off the top of my head think of a writing system produced by a truly wall-less culture, but writing’s only been independently invented a handful of times anyway, most people just crib from another one.

      > Walls have a stabilizing effect, because they prevent wars (Berlin Wall was used as an example, Kennedy – “A hell of a lot better than a war”). If someone’s looking forward to a war, he doesn’t build a wall.
      But that’s exactly what people have done repeatedly throughout history, as mentioned in this very post series! Expecting a war is exactly when people build walls (or renovate existing ones that have fallen into disrepair) to protect themselves. Unless what is meant is an *offensive* war exclusively, though only a fool would fail to take precautions against enemy reprisal.

      I don’t know history well enough to speak to the rest of those claims, but from what little I know these don’t exactly inspire confidence in the author.

      1. >I don’t know history well enough to speak to the rest of those claims, but from what little I know these don’t exactly inspire confidence in the author.

        He’s a prof of history (Eastern Connecticut State University), so he likely knows more about it than both of us.

        I’ve found another article online, it follows the same “look-at-me-I-wrote-a-book” formula.

        “The first border walls aren’t found until the late 2000s B.C., in Mesopotamia. Security is why they were built. There were two different lifestyles developing: a lifestyle of the people I call wallers, who are workers who build things and identify themselves by their civilian occupations. They sought to secure themselves by building structures that would protect them even when they were sleeping at night. Outside the walls, you have a very different sort of society, people inured to the dangers of living in an un-walled world. Non-wallers were peoples we generally refer to historically as barbarians, like the Huns, the Goths, or the Mongols. They were viewed with fear by the wall-builders. And that’s what inspired the construction of the early walls.”

        I think this is his central argument:

        “I would make the case that there would be no writing and nothing as complex as gunpowder without first the construction of walls. The ancient human need for security is one of the fundamentals of life and has to be achieved before we can achieve other things. It was walls that gave people the security to sit and think. It’s hard to imagine a novel being written in a world in which every man is a warrior. Until a society achieves security, it can’t think about anything except the dangers all around it. As a consequence its culture will be limited.”

        “We have this beautiful oration by the Roman writer Aristides, where he is talking about how the empire has become a paradise, how war no longer exists and how Romans no longer believe that war ever existed. They have come to see it as a sort of myth created by old men, while they go about happily to their gymnasiums, theaters, and libraries. It’s all peace within the empire. And Aristides attributed those conditions to the walls that girded it.”
        “The short answer is that, as people become more accustomed to security, they have a lower threshold for what triggers their insecurity.”

        This particular one doesn’t contain a reference to Silk Road though.

    2. Putting any Greek polis in a discussion of fortification would be like stopping an analysis of top-level NFL play to discuss the local state high school team. Putting Sparta in such a series would be stopping that same analysis to discuss a bottom bracket high school junior varsity team.

      The Greeks were really quite bad at siege warfare, both in offense and defense.

      1. I’m not saying this is a great place to discuss Spartan fortifications. But given your opinion on Sparta, I think you missed a good opportunity to criticize them *somewhere*. How boneheaded is that – “I like to fight but I’m too tough to build walls”

    3. Any claim that the Silk Road was protected by a wall is ridiculous based on the sheer length of the thing. The notion that “unwalled civilizations” didn’t produce culture or knowledge is poppy-cock — simply look at OGH’s series on the Dothraki and actual North American native cultures/Central Asian steppe cultures to see some of the art these “unwalled civilizations” produced!

      I can’t find any information on Hadrian II, but my understanding is that the Roman Empire had plenty of walls and fortifications. Hell a Roman Legion built a walled camp every time they stopped for the night!

      I don’t know much about Professor Frye’s methods, but these broad statements don’t lead me to consider his opinions very highly.

  23. “The Silk Road worked because it was protected by a wall.”

    Wut? No, it wasn’t. Maybe for the Chinese portion, for a later part of its history.

    “The North America was inhabited by nomad warrior tribes, and they mostly disappeared. South America was a civilization of wall builders, including the Great Peru Wall.”

    North America was inhabited by many peoples, most of whom had some farming, not nomadism, and they didn’t disappear any more than South American ones did. South America had a lot more *people* and a different colonial pattern.

    “Wall builders accumulated knowledge, produced writing systems, art. Peoples without walls produced military cultures lacking those.”

    The wall-less ‘Minoans’ of Crete produced writing and art.

    Normally a settled city would produce both writing and walls, but it’s not the walls producing writing or art.

    “All great walls required despotic rules”

    The Long Walls of Athens were built by a democracy (as much as any was back then.) If ‘great walls’ means Hadrian or the Great Wall, no one had figured out how to have a democracy/republic *exist* at such a scale to need one — no representative democracy back then. OTOH, is the Maginot Line, built by the French Republic, not a ‘great wall’?

  24. Perhaps not coincidentally, bean’s “Naval Gazing” blog now has an essay suggesting that British ironclad battleship design from 1865-1885 was driven largely by a desire to bring heavy guns to bear against coastal fortifications of that era, rather than for defeating enemy fleets at sea.

    This was never put to the test, at least against contemporary European fortifications, but it does seem plausible. By 1914, improvements in mines, torpedoes, and submarines made close naval attack against fortifications generally impractical even if you could bring heavy enough guns and armor to win the artillery duel.

  25. I last worked on ICBM’s 50 years ago. Started in 1959 as a draftsman making drawings for the 1st stage engine for the Titan 1. One of the interesting projects was to design a way to remove the debris from the top of the MM silos. The silos and missiles can survive to the edge of the ground burst crater. The next problem was how to armor the nose cone so the missile could survive flight thru airborne debris.
    I remember “hall talk” that if we knew the location of the Russia silos we could put a warhead down the hole. That was 50 years ago.
    We need to deactivate the MM systems. Launch on warning is too damned dangerous. Use it or loose it.
    What really scares the hell out of me is the coming of Kitchen Table Laser Isotopic Separation.

  26. “Baen no longer sells post-publication-date ebooks as part of the deal that they made with Amazon to sell ebooks through Amazon instead.”

    What? Plenty of non-bundled ebooks still on Baen.

    1. As far as I know, and I’m a pretty avid Baen fan, the info in this then contemporaneous article is still valid

      Baen sells the monthly bundles up to the cut off dates and eARCs, but nothing once it’s been “published.”

      There was a huge to-do about this at the time, with Toni W being accused of “selling out” Jim Baen’s vision and legacy

      1. Your own article says it’s just about bundles:

        “Future Webscription-style e-book bundles will be still be available for purchase as serialized pre-publications only until the official publication date, after which they become single-purchase-only titles

        The titles are available for purchase from Baen, DRM free. You can go to the site and look, as I did.

        1. I see. Huh.

          The last time I looked (which was a while ago) they had buried the option to buy singles deep enough I couldn’t find it.

  27. The premise everyone post 1940 has been operating on it seems is that anywhere a target is known to be, can be hit by a warhead big enough to destroy it. Suppose a railgun or anti-missile laser in a permanent emplacement became accurate and reliable enough to hit anything you threw at it. So as long as the target was above the horizon, it was vulnerable to this weapon. But it’s large, so it’s not practical to move it (maybe on the largest ships). It takes enough energy to require a dedicated power plant. So then the enemy has to either accept this dome of protection is invulnerable to their missiles, or attack it on the ground. Which will require ground defenses, but ones which are forced to protect the railgun site, not able to freely trade ground for maneuver.
    If such a device existed, it would certainly motivate a lot more fixed fortification because having that dome of protection would be super useful to have for many places.

    1. That was exactly the idea behind the land based railgun efforts. However the hypersonic programs advanced enough that they have surpassed what the railgun could do. Also MDA never fully bought into the railgun and always preferred anti-missile missiles.

      Part of the problem with the railgun was also the team designing it. BAE bought out IAP and had them do most of the design work on the Naval railgun. IAP was a good basic research house but they sucked at applied. They designed the gun to survive things like a stuck armature or a round coming apart in bore. You pay a massive weight penalty for this. In a 6.1 or 6.2 prototype you can deal with it but not with something you intend to field. None of our conventional guns would survive an equivalent condition so why should the railgun?

        1. Hypersonics (within the lower lets say 50km of the atmosphere) have severe issues with manoeuvring and just plain staying in one piece. They do work on paper, as long as you build them out of unobtanium materials that have physical properties bordering on the physically impossible.

          However, should you manage to build such a device (eg one that is both manouverable and able to withstand the massive frictional heat generated), then that material will trivially withstand whatever energy your laser will impart into it.

          If your laser, at the target, is high power enough to actually burn through the heat shield of a hypersonic missile, then the missile is either really close (and most likely on a terminal path), so you might as well just use a good old slug-based CWIS, or your laser source is also on a hypersonic missile, in which case just manouvering close and blowing yourself up is a better idea. Shrapnel does not require perfect, unbroken LOS, a laser does.

          Remember, if you want to use a laser, you need to be able to focus a needlepoint-sized impact area onto a hypersonic target that is flying on a trajectory most not ending in you. The angular velocity of it, and thus your tracking requirements, are insanely high.

          There has been a lot of discussion on these two (and railguns) in the comment section, most of it ignoring physics. One may break the 10 commandments, but the Second Law holds.

          As a collolary: of course there are tech demonstrators that work in idealised use cases in an idealised setting and can be show to get the next 10 years of funding. It is what I would do if I was a military research physicist, and my livelihood depended on selling flashy ideas to military and political personel who could not do an integral if their life depended on it.

          1. Sadly there is no edit function.

            This video by a physicist explains the issues with hypersonic flight (within the atmosphere) pretty well.

            Section on missiles from 11:40

            You need materials that can withstand 2000+ degrees (C or K, at this point it does not matter), many tens if not hundreds of bars of pressure, are still reasonably chemically inert under these conditions, etc, and that is just the hull.

          2. my though on hypersonic in-air missiles is that they have one hell of a signature, and are carrying all that kinetic energy on a predictable track. CIWS doesn’t have the range, and a sea level sensor has too short a horizon, but a USN CVBG isn’t limited to sea level sensors (or a human OODE loop, either). The tech for hypersonic missiles necessarily allows for cheaper, less capable countermissiles to be effective against them with sufficient sensor tech.

  28. This has nothing to do with the current subject but you should really update your “ressources for teachers” and “ressources for world-builders” pages, as there are a lot of missing series.
    It makes it harder to find a specific article when we want to – which is generally when I’m baiting someone to read your whole blog with a tempting first subject;

Leave a Reply