How It Struck a Contemporary: The Armenian Defeat in Nagorno-Karabakh War One Year Later

[Still apparently unable to write anything more deliberate than comments online, I thought I could at least gather some remarks I made in real-time during the Nagorno-Karabakh War of last year. I believe a few of them may have a broader interest than the war itself for people who follow this kind of subject. Lightly edited to avoid (even more) repetition.] 

Azeri success and the psychology of defeat

That the Armenian covering force in the plains to the southeast of Shusha was heavily defeated is not really the mystery. It was a vulnerable position held in insufficient depth by a weaker force. If one adds the devastation caused by Azeri drones in the first days of the war, that this covering force was defeated can hardly be surprising. The puzzle is about what happened next. Put very briefly, Armenian defense in the south concentrated on securing roads leading SW and SE from Shusha, leaving the mass of hills to the immediate south unguarded. And this was where the Azeris made the advance that, ultimately, won the war for them. Quoting now from an even earlier post:

tl/dr, Azeri commandos and light infantry are doing something roughly similar to what Chinese did in Korea against Americans: infiltrate along the heights on either flank to dislocate an enemy that has based its defenses on roadblocks and fortified towns along roads. .. Azeri troops are not holding that central position south of the town without loss, but, on the other hand, there's so far been no sign that the Armenians will be able to push them out anytime soon.

There's really no way to avoid the conclusion that the Azeri ground forces earned this victory: they pushed really hard and suffered the casualties for it. Fairly heavy losses are reported among their mid-level officer corps, with a lot of company and battalion commanders killed, but, within reason, this is often a good sign for combat effectiveness. It's not just that it's good for morale for officers to be visibly sharing the risks of an advance, but it shows that they're close enough to the front to get a sense of the fighting and react to changing circumstances. Those Syrian mercenaries sent by Turkey were handy to reveal positions by having them getting killed in frontal attacks, but there were not nearly enough of them to carry out the offensive. 

I'd be extremely curious to find out to what extent Azeri infiltration across high mountain terrain (as opposed to roads) was actually contested by Armenian infantry. For some reason --could be disorganization, could be fixation with old plan of defense that didn't foresee infantry advance from the south-- there really didn't seem to be that many Armenians contesting the hills paths and peaks. I suspect Armenian command may have deceived itself by rationalizing that with these infiltrations the Azeris were just self-ambushing themselves, creating pockets where they would be pounded by artillery and eventually shot to pieces when they approached defended localities near the roads and towns. 

By the time the scale of the collapse in the south became clear and the Armenian command awoke to the danger of Shusha being enveloped, it seems quite clear that it had entirely lost its grip on the battle. While with perfect information, looking at a map with the actual unit strengths and positions around Shusha in the later stage of the war, it may seem that there was a "favorable operational-tactical situation" for a counter-attack against a strung-out Azeri advance across very harsh terrain, Armenian higher command had nothing of the sort. Last self-quotation from a post written when the thing still-ongoing.

It's one of those cases where maneuver warfare begins to hinge on psychological subtleties. Most penetrations of a front line will be, if you look at it abstractly and with full information, extremely weak and vulnerable relative to the forces available to the defender. It should be possible to organize a counterattack or at least move forces to seal the thing off, but the problem is that defender has only the most tenuous notion of exactly what's happening, he's getting situation reports that are hours or even days out of date, basing counter-moves on assembly areas that have already fallen, etc.., all that stuff fans of John Boyd describe as consequences of getting inside the enemy's "decision loop". 

Thus, instead of problems with "morale" (which I find dubious in general and especially in this case: Armenians appear to have fought hard wherever they could fight), what I'm looking for to understand what happened is (a) Armenian reserves and their positioning and (b) their command and control. The latter factor in particular seems to me to have undeniably played a part in their defeat, they were perpetually on the back foot since their first line in the south fell, their leadership not really giving the impression (to the extent one can judge these things) of having a good grip on what was happening, nor of being able to communicate with itself effectively. Let us look at this factor a little more closely.

Command and control

A breakdown of organization and command and control answers the question of Armenian defeat in a way that raw numbers cannot. Russian intelligence was putting the number of reinforcements/volunteers sent to the place as 20,000, which means that, even with losses, defenders should've had ~40-50,000 people, which is certainly enough for Armenians to have placed some more infantry to hold approaches to Shusha while still defending northern and eastern flanks.

I wonder if a problem was that too little thought had been given as to how to organize and command a defending force that, at the outset of the war, would need to double or triple its own size with all sorts of volunteers and reservists. This is one of those things that sound simple but are really quite technical, tedious, and finicky to do: European armies before WW1 made it look easy, but they spent a lot of time and money making sure this work (part of France's defeat in 1870 was because they tried to improvise this). That, in effect, only a mass of individuals or small units of people could be sent as reinforcements (as opposed to regular units from the Armenian Army) might help explain why the effect was, to quote Lincoln about sending people to McClellan, "like shoveling flies across a barn".

One's eyebrow might certainly be raised at those pictures, genuinely moving on their own, of several generations of Armenians, of both genders, moving up to the front. This is all very well and indicative of genuine motivation at an individual level, but by the time they were that close to the front these people should already have been streamed by a comprehensive rearguard organization into different units or at least organized formations. Here again I wonder if that extraordinary Armenian victory in 1994 wasn't a bad teacher, suggesting that there was nothing Azeris could do that a flood of highly-motivated Armenian volunteers organized into scratch units couldn't defeat. Against a regular army, leavened with a few units of expendable Syrian bashi-bazouks, that fought a very methodical infantry-artillery offensive this kind of improvisation may have been fatal on its own.

Part of the issue was I think a tacit agreement to, by an large, keep the war within the Nagorno-Karabakh area, since technically the Armenian army moving into the place would constitute an invasion of Azeri territory as recognized internationally. But there'd be no real reason why Armenia couldn't have had a covert mobilization plan by which large segments of its army's command and support elements somehow "show up" supporting local militias now mysteriously augmented by large numbers of new "volunteers". We've seen with the pro-Russians in Donbass that this kind of army can be devastatingly effective at modern combined arms warfare. But, for whatever reason, Armenian reinforcements seemed to consist of lots of tiny groups of people moving up to reinforce the frontline. Again it is entirely true that it was with something like this that they won a smashing victory in '94, but this time I suspect it contributed to command and control problems

Looked at abstractly, the calculation was perhaps that if the war could not be won by the existing forces in the province reinforced by hastily-organized militia infantry, then the war for the province was lost and it was not worth having the "army in being" and its equipment and ammunition getting destroyed. A largely-unengaged regular army might be useless for Nagorno-Karabakh, but vital to protect Armenia *proper* from even bolder offensives by genocidally-minded Azerbaijanis fresh from triumph in N-K.  All this said, that supplies of fairly mundane supplies like Soviet-era artillery munitions and even food were collapsing less than a month after the outbreak of the war is still quite extraordinary and unavoidably suggests very poor planning or atrocious execution.

Supply as a signal about leadership quality

There's a rather subtle dimension to the principle of "an army marches on its stomach" that has nothing to do with nutrition. Regular supplies of food are one of the indicators troops have to gauge the quality of their army's command and organization. You could see every individual ration issue as passing a test troops are constantly making, consciously or not, to determine the quality of their leadership. Missed meals, vague or incorrect orders, marches and countermarches leading nowhere, meeting other units that are as confused as yours is--all these things can build up into the impression that your commanders have no idea what they're doing, which in turn worsens the effect of every exertion and loss because of a growing fear that it'll all probably be for nothing.

"Living off the land" is in general one of those things that sound much better in historians' prose than in actual practice. Napoleon got away with it because for a good chunk of his career he was fighting in exceptionally prosperous and productive areas of rural Europe (North Italy and Germany), but the supply habits picked up from there became lethal whenever he moved troops to less hospitable areas in, say, Spain and later Russia (over half the losses of the Grand Armee before Moscow and the Russian winter).

So, back to NK, having troops rely on local villagers giving you sausages is one of the reasons I'm not getting a great impression about Armenian preparations for this war. For prudence or even mere completeness' sake, you'd think that, faced with the prospect of invasion by an enemy superior numerically and materially, their high command would've taken care to stock plenty of supply caches throughout the place (i.e, not just in the frontline), but nothing like this seems to have been done. (Or, perhaps, it was done at some point but the stuff then either stolen or sold off: I gather that within Armenia and the diaspora there's a lot of talk about corruption having crippled their defense.)

The effect of supply breakdowns in undermining the confidence of soldiers in their leaders is one of the reasons why the morale effect of casualties can seldom be stated as a percentage, i.e. "units become combat ineffective after X-percent losses". Let's look at a hypothetical example from this latest war based on what we've learned from several post-mortems. Azeri units at the point of the spear in the attack to the south suffered very heavy casualties indeed, including many of their commanding officers, but they were part of a victorious advance that was visibly approaching an objective of high strategic and symbolic importance. A defending Armenian militiaman will have been hungry ever since consuming whatever food was packed for him at Yerevan, has been dicked around by multiple commanders giving contradictory orders ever since arriving in theater, saw units advance and retreat with no clear pattern, and is beginning to smell the panic that can spread in areas behind the front when the higher command is visibly losing its grip of the battle. Thus, a loss of 100 men in the Azeri unit will arguably have much less effect than the killing of one guy in the Armenian: the former will be processed as the accepted price of a demonstrably approaching victory, while the latter will likely feel like an utterly pointless death that should encouraging everyone else not to be the last person to die in a fucked-up campaign. 

Troop quality and "good fighters"

"Troop quality" is a bit like casualties or "morale" in that it's stuff that's considerably easier for historians to write about a decade or so after the fact than it is useful for we average war nerds trying to make sense of a war in real time. These are variables that only start to have an independent effect when they reach truly extreme levels (my econometrics too rusty to remember how this'd look like as a regression).

So, for example, the Saudi armed forces really are so spectacularly awful that you can say that a given engagement was probably lost because of low troop quality, the effect of this variable swamping any other advantage they may have had (such as Abrams tanks). However, with almost any other army, even extremely mediocre ones, it's much safer bet to focus on stuff like training, personnel-stability or equipment levels, because an army you may have dismissed as shitty from one set of engagements may turn around its performance much quicker than you'd expect if you think the latter was explained by some innate quality as fighters. So, for a last example, compare the Turkish Army in northern Syria in 2015-16 vs Turkish Army in Afrin. Latter not a glorious triumph of the operational art, but whatever defects in intrinsic "quality" its troops may continue to have was more than offset by intelligent tactics applied professionally by a reasonably straightforward leadership structure. And these problems with troop quality arguments turn out to be true for even the most commonly known examples of "bad" armies, e.g. the WW1 Austrians or the WW2 Italians--under the right circumstances, a properly prepared unit that was well-commanded would do fine.

The problem is really developing a general idea that a certain army or people are "good fighters" without paying attention to, for example, how much and for what they're trained. In the case of Armenia, a lot of people (including, most damagingly, many among the Armenian command) seemed to have drawn from the 90s the lesson that Armenians are invincible mountain fighters whereas Azeris are plains peasants who start fainting when the elevation goes up too much.

There's a slightly more subtle point here that I think can be missed sometimes even by professionals. Wars tend to be won not so much by men as by teams, so it's usually better to have ten mediocre soldiers who've trained together for a long time rather than ten black-belt super-duper operator commandos who've only just met. So far, so very platitudinous, but less so is the observation that this principle continues to hold as you move up an order of battle. Thus, better to have four mediocre battalions that are used to working together as a brigade than ten excellent ones that you've only just "task-organized" into a battle group. A problem many modern militaries have is that maneuvers and exercises by things like divisions or even brigades are very expensive and a great deal of trouble for everyone involved, so they tend to be skimped on in preference for training at the level of battalions or even companies. The quality of these lower level units may indeed be quite good, but this may not matter if they're thrown into actions where they have to work as part of a larger formation.

So, back to Nagorno-Karabakh, even if every one of the Armenian reservist and volunteer companies were elite and highly motivated mountain fighters, the problems coordinating their activities over a changing frontline may have meant that their performance was collectively inferior than that of an Azeri army that seems pretty much to have fought the war it trained for.

Coda

Defeat is often the best handmaiden to historiography, so look I forward to the Armenian post-mortems.