Thursday, August 7, 2014

What Is Goodwood To Us, Or We to Goodwood? Totalising Change

Terry Copp summarises: "After a successful assault on a defended coast, General Eisenhower's naval, air and ground forces destroyed two powerful German armies in just seventy-six days. Enemy losses of close to half a million men included the combat elements of thirty-seven divisions deployed in Normandy as well as another six left behind to delay Allied access to ports. Other large fragments of Army Group B --about 25,000 men from twenty different divisions-- were encircled at Mons on 4 September." Allied operational researchers counted more than 8000 damaged, destroyed or abandoned vehicles, including 456 tanks and self-propelled guns and 367 other AFVs. An estimate of uncounted vehicles suggests that the actual total was 12,000. Twenty thousand vehicles, including 250 tanks and self-propelled guns, escaped across the Seine to continue the fight."

For those inclined to think graphically, we have Terry Copp's research in the OR group's files to thank.

(Fields of Fire, 251).

That's one sense of a "Grand Finale," although the one embedded above for the sake of the thumbnail is of the end of a shindig thrown for the Canadian gunners by the highways folks. The Gunners have a sweet deal with the Federal Highways department where they send guns up into the passes in the winter to shoot at snow. The idea is to start an avalanche before the snow pack builds up to the point where a slip can climb up the other side of the valley and bury the road.

Sometimes, the deal means that you get a party thrown in your honour. Other times, it means setting up a gun position in Rogers Pass in February, which is not as nice as you'd think. 

The metaphor here gets highways, mountains and guns in it, but I am leading with it because it is an anthropogenic intervention that transforms the steady state of a resting snow slope into a chaotic transformation-of-state. The gunners choose to fire the shell. What happens to, say, a log caught in the avalanche a moment later is utterly beyond their control. In this comparison, "log" means just about everyone and everything involved in World War II. Wikipedia's "Battle for Caen"  framing article lists eight named, corps-level operations between PERCH (9 June) and  GOODWOOD (18-20 July), better than a major operation a week. With the weight of 2nd British Army shifted west for BLUECOAT and a first, jarring encounter with the Jagdpanther (sometimes confused with the Jagdtiger (link; better link), 1st Canadian Army was left to fight SPRING (25 July).  intended to capture the social changes enacted by World War II. They were not chosen; indeed, were beyond intervention. All one could do was fire the shell, and see what happened. The metaphor also serves because it has semis and guns in it. The great changes launched byand coming to their culmination with TOTALISE (7 August), and TRACTABLE (14 August).

Put this on a calendar. Imagine it in your life. There is a two week delay between SPRING and TOTALISE. That is very, unusually long by the standards of planning and preparation for one of the major Normandy operations. It was long because it allowed the Canadians to reconstitute the abused infantry battalions of 2nd and 3rd Canadian infantry divisions and establish 4th Canadian Armoured and 1st Polish Armoured Division in the staging areas freed by GOODWOOD. The length suggests an usually well-planned operation. GOODWOOD, recall, matured over about 10 days. Fourteen implies another four to think and prepare! But, again, put this in perspective of your own life and remember that it is nineteen days from GOODWOOD to TOTALISE. Do not expect revolutions in human thought in that length of time. There is change here: great change. But it is the change of the avalanche, carring all before it.

Etc. Etc: I do not mean to be vague here. The change I am thinking about is very specific. The men who fought TOTALISE came from farms where horses still pulled things. They went home and erected the awesome (and awesomely ugly) reinforced concrete pillars that allowed the original Park Royal shopping mall of my childhood to park cars on its roof less than thirty years later. This is our modern built environment. At its best, at least, the brightly lit urban caves that you guide your vehicle through in Lego Racers. (Not that I could find footage of those levels on Youtube, but you know what I'm talking about. I hope.) Our idealised modernity is a a LEGO world made in 100 ton blocks. 

First, though, guns. Here is a very, very ugly one, a United States Army 105mm Howitzer Motor Carriage M7. A very quickly reworked variant of the M3 Medium Tank. Interestingly, unlike the tanks built in the state arsenals by GM and Chrysler, this machine was laid off to the locomotive industry entirely, and built mainly by diesel pioneer Alco (American Locomotive Company). It looks like the Ordnance Coprs decided that it would be a good idea in late 1941, that the  lead contractor's idea of what it would look like was approved for production in February of 1942, that the British Tank Mission ordered 5500 in the spring, that no more than 4000 were ultimately delivered, mostly to the United States Army. It was by all accounts a perfectly satisfactory weapon, given its limits, but a look at the thing will suggest the nature of those limits.

The look will be familiar, and "refined" is not the word. The chosen engine was Continental's version of the venerable Wright Whirlwind in its nine-cylinder version. That's a radial engine. In an armoured fighting vehicle. A radial engine is defined as (approximately) opposed cylinders in a radial configuration, hence the name.It is really, really tall (and wide) by design

The problem here is that it is an armoured fighting vehicle, which is also a gun. That means that it starts out with a pretty major strike against it. The gunners of 1942 were brought up in a milieu in which horses drew guns, and in which horses were easily and routinely switched out. They're valuable, and they go lame. So one halts, hitch in another.

Beyond that, guns don't typically look like the above. Here is another gun

As you can see, the focus of the engineering is very much on making sure that the muzzle of the gun can be pointed to low and high elevations. (-5 to 45 degrees in this case). That's so you can put it on a hill and not be worried about the  enemy in the valley below, and, conversely, shoot a very long distance. In this case, an 82lb round to 18,200 yards. Put a gun on a chassis that can't do that, and you've built a pretty useless gun. Put a gun on an armoured fighting vehicle that can do that, and the case can be made that you've built a somewhat-less-than-useful AFV. 

To be as clear as I can be here, armoured fighting vehicles, and guns, are supposed to be low to the ground. Putting a radial engine in one is stupid. It's productively stupid, in that it gets the "Priest" into action. It's still in the spirit of desperate improvisation. That is how the avalanche works. One expedient after another takes you in directions you could not have mapped out if you had all the time in the world, never mind two weeks, at best.

Now, to set the scene: we have seen that, whatever his original plans or brief excursions from them, the overarching objective of the Normandy campaign is for the British and Canadian troops, operating on the Paris side of the beach-head, will maintain pressure on the Germans. The Germans, understandably, will not pass many reinforcements down a line parallel to the front. And so the "colossal cracks" keep coming:  EPSOM (26 June), CHARNWOOD (9 July), GOODWOOD (18 July), BLUECOAT (30 July), with smaller, corps-level operations in the midst. 

Far more importantly, there was the American breakout from St. Lo. COBRA, in planning for over a week, was a fairly obvious next step for the Americans. Here's the official history's map, borrowed from Wikipedia:

And here's a Google Maps capture, centred on Argentan and blown up to give you the American staging port at Dartmouth, Exeter, and Paris:

Orient the maps and you will see that St. Lo is the classic tabletop wargamer's strategy. By hugging the map edge, the Americans cover one flank; by cutting Brittany off at the base of the peninsula, they instantly solve the Allies' basing area problem by giving them all the space they need for ammo dumps and workshops. And they get astride the Brest-Rennes-Le Mans-Paris axis, opening up a second approach to Paris.

What happened after that, of course, is that the German position in the south collapsed. With their forces concentrated around Caen, they simply could not hold the skillful American attack. The collapse signalled what El Alamein and the 1943 retreats in the East had promised: the Germans had adopted a basically static defensive doctrine that turned on weapons like the 10/15/32cm rocket mortar launcher, 88mm AA/AT gun, and the Panzer VI: weapons that the Germans simply could not maneouvre in the sense of rapidly shifting from sector to sector with the available stock of wheeled transport and prime movers. The Germans had built a train-mobile army, and the trains were not available. 

COBRA signalled that the time had come to punish them for that. The ambitious General Guy Simonds, repeatedly promoted to become perhaps the least experienced corps commander in the Commonwealth force engaged in Normandy, was told to do just that. Simonds' rapid promotion can be laid off on his imagination and ability. Summarising TOTALISE, wikipedia has it that: 

For Totalize (beginning August 7), which involved a night attack, numerous navigation aids were devised, along with heavy bomber support. Having learned from Operation Spring, Simonds devised the "Kangaroo", an early armoured personnel carrier converted from non-operational armoured vehicles "defrocked Priests".[1][24] Granatstein characterizes the plan as "brilliant if too complicated",[25] in that it did not account for the inexperience of the troops.

Here, at last, I circle around to my point in taking random potshots at Alco's best. The implication that people writing about TOTALISE are striving to make without clearly defending is something along the lines of "Simonds invented the Armoured Personnel Carrier." A quick review of the wiki article (again) does not even begin to get at just how wrong this is. In later years, Basil Liddell Hart would congratulate his bestie, General Fuller for being the lone prophet of the "tank marine," and the source is enough to tell you the real story so that you can turn to the 1925 number of the Army Quarterly and read Bernard Stewart Prize Essay contestant after contestant (that is, up-and-coming British army officer) arguing that the future lay with the APC, but that the present did not, due to costs. The first cut of the difference between 1925 and 1944 is that, for reasons that must have seemed good at the time, 3rd Canadian Infantry Division landed at JUNO Beach equipped with M7s in its field artillery regiments. Come August 1st, for reasons that also seemed good at the time, it was in the process of being re-equipped with towed 25 pounders. That made the M7s redundant, and available for other uses. "Defrocked" tanks and other AFVs were popping up on every front and in all sorts of roles. One of the units that counterattacked TOTALISE would be a German unit of self-propelled guns improvised out of turretless old, French tanks. Junkyard wars was well on.

Now, roll the clock back for a moment to GOODWOOD. The controversy over GOODWOOD is that it was sold as a "breakthrough," and turned  into a "break-in," or "bite-and-hold" operation: the WWI conceptual precursor to the "Colossal Crack." So let us, by all means, turn the clock back to World War I and the key problem that led Allied commanders of 1914--18 to abandon the "break through."

It was, simply, this: 
At a little over 5 tons for the barrel and cradle only, the 9.2" breechloading howitzer was the descendant of the classic "artillerie du place," the big siege guns that were brought up to the outer defences of a fortress via the native heavy-load moving capability of the existing civilian economy, erected as a working machine at the site, and used to smash concrete, masonry, and make whatever was solid that military (and other) architecture could devise melt into air. A 300lb shell, thrown a maximum of less than 12,000 yards, or 14,000 yards in its final war-era iteration seemed more than enough.

And then war came, and it was reinvented as the Commonwealth's main "counter battery weapon." Let's recall that the 5.5" gun-howitzer, shown above, went to war with a 4.5" partner intended specifically for counterbattery work

Under six tons at the towbar, ranging out to 20,500 yards, but throwing a mere 60lb round. The Medium Gun is admittedly a bit of a war orphan, its role eventually usurped by the 5.5" firing a streamlined shell, but the point here is the performance: range, not weight. Because you cannot "break through" an enemy position unless you can suppress their defensive fire. And you cannot suppress the defences unless you can hit them. And you cannot hit them unless you can range them.

Why the sneer italics? Because recall  the essence of the new German defensive position. Light AT forward, heavy AT forward, creating interlocking zones of denial of observation. Hidden fire positions that engage the enemy infantry as they come forward. Admittedly, they are no longer hidden by natural hillocks and swales but by Norman orchards and n farm compounds, but they are there. Once the Allied infantry penetrates beyond direct observation, onto the dreaded "reverse slope," they will be attrited by mortar fire (rocket assisted or not), counter-attacked by "penny packets" of armour. (It turns out that penny packet tank tactics are awesome when the SS do it.)  GOODWOOD is an all-tanks operation. It depends on the suppression of German defences in depth. Everything that can be suppressed can be overrun. GOODWOOD makes no sense, in terms of detailed, drill-down tactics, unless the German defences are attacked and suppressed in their full depth, allowing the tanks to break through. 

We know, of course, that GOODWOOD did not work out that way. Looking back, again, at the gap between conception and execution of only ten days, I cannot stop being amazed. The plan puts three armoured divisions across a river into a narrow bridgehead, ideally in secrecy, and then throws them into the bottom of the sack. Those are not sneer italics. That's an invitation to wonder. The traffic control issues are unimaginable. The confidence of the planning staff that they will be worked out is amazing. The accomplishment, of course, falls a little short of of the ideal, in that 7th Armoured Division ends up tagging along behind the Guards and especially 11th Armoured Division, but it is still, recognisably, an attack, and not the traffic-jam-in-empty-space that was the pursuit at the end of El Alamein. 

GOODWOOD failed, Simonds thought, because armour-infantry cooperation was inadequate. The task was left to the accompanying motor infantry battalions of the armoured divisions more-or-less because it was not considered possible, or even imaginable to mix marching infantry in with all of that armour. Nineteen days left him, and the Canadian 1st Army with enough time to think through what had happened, sort out what went right, and come up with a new plan that fixed GOODWOOD's problems.

The solution? Leave out the third armoured division, and replace it with infantry in vehicles that could keep up with tanks. The staff looked around for the vehicles that would allow an infantry brigade  to ride with the tanks. Simonds' plan called for two armoured columns to advance along two routes on either side of the Caen-Falaise highway. Two troops of Shermans in the "gapping force" would cover flail tanks and AVREs. Next would come more troops of Shermans --the assault force-- followed by infantry in battefield-mobile vehicles, followed by all of their necessary mortals, AT guns, and medium machine guns, all towed by armoured vehicles if necessary. 

Crucially, instead of supporting the advancing columns with field artillery, the 5.5" gun-howitzers of the army-level medium artillery regiments would move with the columns, extending the advancing armour's envelope of fire from the 13,000 yards of the 25 pounder to the 18,000 yards of the bigger gun. Ironically, all of the armour made up for the lack of the equipment so recently considered revolutionary and vital for aggressive artillery tactics, the gunshield.  

Out of the necessity for enough armoured carriers to support this ambitious plan comes the 31 July directive to the Deputy Director of Mechanical Engineering of the Canadian 1st Army, Brigadier G. M. Grant, to convert the available M7s into additional carriers. Han Spoelstra has tracked down an account of the work, but not, unfortunately, of the later life and career of Brigadier Grant. (Google does tell me that he was a Trinity College grad, so he had certain obstacles to overcome in life, and perhaps they overcame him.

In the evening of 31 July ...G.M. Grant, was instructed to set up an organization to convert 72 Priest SP Guns, to APCs by 9 August. The date, however, was soon changed to 6 August with 'as many as possible'. The job consisted of removing the 105 mm guns with mount and mantlet, welding armour plate across the gap left, overhauling the radial engine (100 hour check), transmission, controlled differential, brake linings and running gear (the overhaul itself would normally take seven days to complete).  An AWD (Army Workshops Detachment), code named 'Kangaroo', was set up in two fields near Bayeux with the camp sited in the protection of a neighbouring orchard, about twenty miles from the start line. Fourteen Canadian and British units contributed and pooled their efforts and skills in the project. The unit was led by Major G.A. Wiggan and totalled 250 men. The first arrived at the Kangaroo site late on the afternoon of the 2nd of August and had 14 equipments stripped before dark that night. The hours of work were 0500 to 1100, 1230 to 1700 and 1800 to 2200. But as Cfn. A.M. Campbell of 2 Tank Troops Workshop later recalled, he worked steadily for the four days from 0400 to 2300. He was so busy he didn't even know what the next fellow was doing. Each had his assigned job, for example, Cfn. Campbell was track tightening or engine changing using his Diamond-T wrecker.  All the RCEME/REME units were canvassed for the electric and gas welding equipment with the tradesmen to go with them. Almost unlimited access to welding rod, armour plate, radial engine parts, oxy-acetylene welding sets and gases and radial engine overhaul stands was required.  The armour plate came from the Help-Yourself-Park of "W" crocks (those tanks declared beyond repair). After this source ran out, mild steel from the steel mills in the south of Caen was tried. The Navy also complained at this time that Canadian soldiers were cutting pieces of plating out of craft stranded on the beaches. Maj. Wiggan recorded in his war diary that 'An examination of steel plate in wrecked naval vessels and from the Schneider Steel Works in Caen showed that metal to be unsuitable.' Hence it was necessary to fall back on mild steel plate welded over the openings in the form of spaced armour, that is, one plate welded over the opening from the outside and a second plate welded over the opening from the inside. The gaps between the plates were filled with sand. All carriers were completed on time by 2000 hours on the 5th. In addition, 6 more were ready by noon on the 6th. LGen H.D.G. Crerar, GOC 1st Canadian Army, telephoned his thanks and congratulations to Brig. Grant in the afternoon of 5th and asked him to convey these to all the officers and men responsible for doing such a splendid job in so short time. 
Junkyard wars! The point, then, was not that it was a plan to use armoured personnel carriers. It was a multi-stage plan to overcome a multi-line German defence in which the positions to be attacked by the infantry were well ahead of the start line. The infantry needed battle taxis. The REME and RCEME gave it to them. That part worked. It was the armoured fighting, more-or-less, that fell apart. In 2012, the journal Canadian Military History was kind enough to put the relevant war diary of 1st Polish Armoured Division online. It does not really explain why the Poles did not press on to Falaise, but it does strongly imply that they had little faith in their tanks in the face of Meye's counterattack.

It is at this point of interest to extract one lesson about AFVs-in-motion that is not usually taken from TOTALISE, which is that had the original production schedule been met, the Polish reconnaissance regiment would have been mounted in Comets, not Cromwells. Would it have made a difference? It is hard to say with any certainty. Their absence, though, is a pretty stingining indictment of the whole idea that a fixed, "good enough" tank design would be enough to see out the war. The lesson might be that you are in an armour-versus-guns race to win, or that morale is important.  

Or that you can never stop innovating. I like that one. This entry is a meditation --again-- on "failing forward." Every operation on the road from Caen to the "closing of the Falaise Gap" is a failure in terms of the the boldest objectives set for it. Each failure, however, sets the terms for the next. The end of it is, inevitably, the rejection of the idea that you can unilaterally proclaim the end of the agonistic cycle of competition: that you can save your resources and stop innovating. 

Now, finally, to take this in one further direction. I grew up in a little town called Port Alice:

I found this image noodling around the net, and, one more time, here's Wikipedia, with a little gem of a picture of the townsite when it was south of this picture, at the millsite, a little closer to the war years, in 1950.

This is the town when it did not even have a road connection with the outside world, a species of the primordially Canadian. The reason for appending this? Because it is a very small town in a fairly small (population) wise, country. Perhaps 500 people live down in that little town, and we shall not inquire too closely as to just who they were. This is roughly one person for every 24,000 Canadians, which means that, by a strict rule of proportion, twelve young men of this town are in Normandy right now. The 250 men of RCEME Depot "Kangaroo" are, of course, a very small proportion of the total of 300,000 Canadians in the beachhead, but the point here is that there are only going to be 1600 Canadians riding the things into battle, and a great many of them are driver-mechanics. (Quite possibly newly minted ones.) Sure, the unrecognisable names on our quaint little village cenotaph  were more likely to have been infantrymen than welders, but the men who returned to this little town were more likely to have been mechanics than riflemen. 

Here's Google Maps one more time:

Port Alice exists to house mill workers. Since four years after I was born, those millworkers have lived in the village site imaged in the first picture above, which you can make out at the top of the Google Maps image here. It is eight miles north of the mill, far too great a distance to walk to work. But the thing is, everyone has cars these days. (Or, more likely, trucks.) They need them to drive across island, following the incredibly cheaply-built road that probably threads an ancient trading trail through the mountainous landscape, built by the Province because it was unthinkable that a town of 800, and the mill they work at, should be cut off from the modern road network. This in a region the size of Scotland with a population in the range of 10,000 at its peak. 

It really is a new world. Let's at least meditate on the possibility that it was not an inevitable given of the onward march of technology, but rather a product of the agonistic struggle in Normandy in the summer of 1944. 


  1. I don't think a pure skills transfer model works; innovation is, after all, what happens when technology gets out of the control of the upper classes, who hate horrible surprises.

    There was a whole generation where the guys who'd been privates and corporals and sergeants and turned "get this done" directives from staff organizations into success, and knew that they'd done that, and even the staff organizations would admit it on honest days. The present days have the capacity for innovation back under social control, which is the same as not much happening, but you couldn't make any of the arguments for that in 1950 because there were a lot of veterans who would laugh at you. They'd beaten armies by doing what you wanted to claim was impossible.

  2. The key here is the "out of control" part, I think. You need to win the war; you need to do this and this and this. So you throw resources at your working groups and they put together a plan and you look at it, and maybe if you had the time, you'd ask yourself, "What are the downstream consequences?"

    But there is no time. Instead, you say, "Go ahead." It probably doesn't work --but it does put you in a better position, so you call it a partial success, and, really, that's all the time you have for it, because the next problem is already pressing. What comes out at the end is skills transfer, but also new skills --some hard to recognise because we do not have names for them.

    "The computer" is our modern paradigm for that, the most important change to come out of World War II is the emergence of a name for an industry that is already happening. Neuman might give us a name and teach a chourse, but in the meantime a machinist in Los Angeles is adding a card punch to his new rivet sorter so that he can have a running count of output. The machinist doesn't know that he's working on a "computer," and the idea that he is going to escape the working class and become an "information technician" is as much about containing the social rupture as anything else.

    The heavy duty mechanics of RCEME Depot "Kangaroo" are another matter. No-one is ever going to call them white collar workers. They're going to be "general contractors," and buildmalls and the Gardiner Expressway and houses.

    But they are going to get rich and own nice houses in Mississauga.