Azerbaijan’s Victory: Initial Thoughts and Observations (and Caveats for the ‘Innovative’)

Victory in the 21st century

Autumn 2020 saw the Republic of Azerbaijan, a country in the South Caucasus, population around 10 million and annual military budget of just $2 billion, achieve something the USA and UK had not for nearly thirty years – winning a war. 10 November brought an armistice, brokered by Russia, ending the six week-long Azerbaijani offensive into Nagorno-Karabakh – the region of southwestern Azerbaijan which, according to viewpoint, was under illegal occupation by Armenia since the previous Armenia-Azerbaijan war of 1992-1994 or struggling for independence as the Armenian-majority ‘Republic of Artsakh’ – and the seven Azerbaijani districts also under Armenian occupation since 1994 surrounding it.[i] Under the terms of the armistice, Azerbaijan kept the territories it reconquered – four of the seven districts – while the Armenians ceded the other three, Azerbaijan thereby regaining around two-thirds of the territory lost in the 1990s.[ii] To Azerbaijanis[iii], ‘The Patriotic War’ is a historic triumph, the healing of a quarter-century old ‘bleeding wound’ central to their national identity, peaked by the recapture of the major historical and cultural centre of Shusha the week before the peace deal was signed.[iv] For Armenia it is the diametric opposite: the Armenian government and the armed forces of ‘Artsakh’ – the Nagorno-Karabakh Defence Force (NKDF) – failed demonstrably on every level despite obdurate fighting from soldiers on the ground and in the weeks after the peace deal mass demonstrations demanding the resignation of the government took place in Yerevan, some violent, and the period since February 2021 has brought persistent rumours of impending military coups in Armenia.[v]

Why does this matter beyond Baku and Yerevan? To begin with, these events demonstrate there might still be such a thing as ‘victory’ in 21st-century war and a place for conventional military force (i.e., force aimed at contesting territory and delivered by regular armies and air forces) in securing it. This flies in the face of some highly-publicized arguments that thanks to two decades of ‘unprecedented’ cultural and technological developments, ‘conventional warfare is dead’, military confrontations revolving increasingly around non-kinetic assets aiming at largely non-kinetic effects through applying or countering ‘hybridity’ somewhere in the ‘grey zone’, meaning ‘legacy’ capabilities including heavy armour and conventional manned aircraft should go firmly into the dustbin of history.[vi] However, most Western military punditry has focused on the most reported-on aspect of the war (at least outside the two belligerents) Azerbaijan’s extensive use of mainly Turkish-made Unmanned Combat Aerial Vehicles (UCAVs) and its apparent implications for future warfare and current acquisition policy. This includes some influential voices, at least in the UK.

‘Political Geography Now: Nagorno-Karabakh Control Map & Timeline: Artsakh Withdrawals – December 1, 2020’

Writing on potential futures for post-Brexit British armed forces, the Conservative MP and defence pundit, Bob Seely, commented on ‘how Azerbaijan’s use of inexpensive Turkish drones [sic] has decimated expensive Armenian equipment such as tanks and armoured vehicles, changing the balance of power on the battlefield’.[vii] This view is shared, apparently, by the most senior British decision-makers, the Defence Secretary, Ben Wallace, describing Turkey’s use of ‘drones’ in Libya, Syria and ‘elsewhere’ as ‘game changing’ and stating that the UK had some catching up to do, with implications for the upcoming UK Strategic Defence Review, rumours abounding of troop numbers and ‘legacy’ capabilities being cut in favour of greater ‘automation’ and ‘innovation’; it was reported subsequently that the review would commit to purchase cheap ‘attack drones’ based directly on evidence from Nagorno-Karabakh.[viii] The former chair of the House of Commons Foreign Affairs Committee, Sir Richard Ottaway, was more nuanced, acknowledging the part played by Azerbaijan’s careful alliance-building and acquisition policy but also claiming that the UCAVS were ‘pivotal’ in assaults on Armenian defensive positions and arguing the UK should develop its own low-cost equivalents to the Turkish TB2s employed in Nagorno-Karabakh to supplement the (highly expensive) Predators supplied by the US.[ix] No such nuance across the Atlantic, predictably, an article in Forbes, for instance, hailing the TB2s as a ‘silver bullet’, alluding to suggestions from some quarters that ‘the massacre [sic] of Armenian armor signals the end of the tank’ and that the acquisition of cheap UCAVs by small powers might provide challenges even to US forces; Radio Free Europe was even more effusive, proclaiming that ‘In Nagorno-Karabakh, the Future of Warfare is Now’.[x]

This is the author’s own contribution to this debate – his observations on what happened in Nagorno-Karabakh in autumn 2020, some suggesting many of the claims cited above may need to be nuanced or revised. He begins with strategy and the role of conventional warfare in achieving Azerbaijan’s policy aims.

Strategy and the Utility of Force – Conventional Warfare is alive and kicking Armenia

While the war demonstrates that conventional military force can still be brutally effective in securing policy aims, those aims must be clearly enunciated with an obvious and realistic political end state in mind and enjoy strong public support – things we have emphatically not seen in the West since the mid-2000s. It also helps if those aims can be framed in terms of securing objectives of geopolitical significance – something which conventional ground forces are designed to do – rather than more diffuse ones of ‘fighting terror’ or ‘implanting democracy’. While popular at home – even more so, now – President Ilham Aliyev is viewed widely outside Azerbaijan as an iron-fisted autocrat with an indifferent human rights record and allegations of industrial-scale corruption of Western legislators and officials laid against him.[xi] However, in 2020, President Aliyev (and given the highly personal nature of his rule, we presume it was mainly him) formulated realistic policy aims attainable by the means available centring on seizing key ground, as he made clear in an interview with the BBC on 9 November: if the Armenians committed to withdraw from the seven occupied regions around Nagorno-Karabakh (four largely re-taken by Azerbaijani forces by then) then he would halt the offensive and be willing to negotiate the future of Nagorno-Karabakh.[xii] However, he promised he would ‘fight to the end’ if they didn’t withdraw and also demanded a right of return for the Internally Displaced Persons (IDPs), Azeri families who fled to Azerbaijan from the occupied territories during and after the 1994 war, their numbers estimated by the United Nations High Council for Refugees as a possible 1.2 million by 2014.[xiii] Compare this with the outcome: President Putin’s peace deal mandated that Azerbaijan keep the four districts it had re-conquered while the Armenians handed over the whole of the remaining Aghdam, Kalbajar and Lachin districts at the beginning of December – so the war clearly achieved Aliyev’s primary geopolitical aim while opening paths for the others.[xiv]

He had also strengthened Azerbaijan’s hand via cultivating the right allies and if anything symbolised this, it is President Erdogan of Turkey being guest of honour at the victory parade in Baku on 10 December, a Turkish Army contingent marching past he and President Aliyev to the strains of the Ottoman march, Cedin Dedden.[xv] President Aliyev’s father, Heydar Aliyev, founder of modern Azerbaijan, described the relationship between the two countries as ‘One Nation, Two States’ – two Turkic peoples, with common ancestry and culture, speaking mutually-intelligible languages, with traditional enemies in common and therefore natural allies, an assumption shaping the younger Aliyev’s external policy and suiting President Erdogan’s ambitions for his country, also, Azerbaijan now providing a powerful bridgehead for Turkish influence in the Caucasus and Central Asia.[xvi] The two countries cooperate closely in the export of oil and gas dominating Azerbaijan’s economy and shaping the politics and society of this classic ‘rentier’ state, most prominently on the Baku-Tbilisi-Ceyhan oil pipeline through which up to 9 billion barrels per year flow on their way from Azerbaijan’s oil fields in the Caspian Sea to Western Europe.[xvii] Military cooperation is just as close, giving Azerbaijan the benefit of learning from a respected NATO military with credible warfighting capability tested for real in Syria and Libya. Turkey has provided Azerbaijan with training support and military equipment since 1992, Azerbaijani cadets attend the Turkish Military Academy at Ankara and the two armies exercise together frequently. In 2010 Turkey and Azerbaijan signed a treaty in which each promised to come to the aid of the other if attacked: a formal alliance in all but name, and an obvious expression of this was Azerbaijan spending unspecified millions – out of that defence budget of $2 billion per year – on buying 50 TB2 Bayraktar UCAVs from Turkey along with MAM-L laser-guided bombs (also Turkish made), the main weapons used by the TB2 in the war. Israel, another close security partner, provided Hermes and Heron reconnaissance UAVs while Azerbaijan’s own Azad system was also used for reconnaissance. Israel also supplied Harop loitering munitions, which Azerbaijan first used during previous fighting along the line of control with Nagorno-Karabakh in 2016. Given the TB2s arrived just a few months before the offensive began, it is entirely possible that some were flown by Turkish Air Force pilots, possibly from inside Turkey itself.[xviii] Another, just as telling expression of the alliance was the thirteen-day long exercise in Azerbaijan in August 2020 in which up to 11,000 Turkish troops participated and saw Turkish and Azerbaijani Special Forces practicing airmobile assaults on high value targets alongside each other.[xix] This close relationship paid off in fighting the war.

Fighting the war – was it the ‘drones’?

The war presents a real-world example of Clausewitz’s concept of strategy as ‘the use of engagements for the object of the war’, particularly as the very object of the war was seizing and securing territory, achieved through a series of battles.[xx] President Aliyev was smart enough to leave the management of this fighting to others: planning was carried out by the General Staff under Colonel General Najmedin Sadykov, like all senior Azerbaijani officers a product of the Soviet military education system, and a solid grasp of operational art is evident throughout.[xxi] The offensive was carried out by 1, 2 and 3 Corps of the Azerbaijan Army – a force of sixteen motor-rifle brigades with two artillery brigades controlled centrally – and consisted largely of a methodical advance aimed at seizing key towns, villages and chokepoints, resolving into two broad foci of effort: a push in the northwest of Nagorno-Karabakh aimed directly at its capital, Stepanakert, resolving itself into positional fighting around the town, going alongside an advance in the south through the more open country of the occupied Azerbaijani districts of Fuzuli, Cebrayil and Zangilan, aimed at the town of Zangilan in the far southwest – taken on 20 October – securing the entire length of Azerbaijan’s border with Iran. The taking of Zangilan was followed by a renewed offensive northward leading to the taking of Shusha, the centre of gravity of the whole of Nagorno-Karabakh: not only does Shusha have enormous cultural and emotional significance for both sides but it dominates the Lachin Corridor, the only line of communication between Stepanakert and Armenia and so taking it rendered the Armenian position across the whole of Nagorno-Karabakh untenable.[xxii]

Armenian/NKDF forces based their strategy on defence in depth with the apparent aim of making Azerbaijan’s advance as costly as possible. Nagorno-Karabakh’s mountainous terrain made such defences viable but the mountains also meant that lines of supply were limited and control of certain chokepoints was key – hence the importance of the Lachin Corridor.[xxiii] This went alongside using artillery, a combination of BM-30 Smerch multiple rocket launchers and Scud-B and SS-21 ballistic missiles, to attack cities in Azerbaijan with the aim, according to source, of attacking Azerbaijan’s deep lines of communication or of terrorising the population and putting the Aliyev regime under political pressure. This began on 4 October with missiles fired at Terter, Mingachevir and then against Ganja, Azerbaijan’s second largest city and nearly fifty miles behind the front.[xxiv] Ganja was hit again on 5-8 October, 11 October and 17 October each time killing civilians (for balance, Azerbaijani forces barraged built-up areas with Smerches, sometimes with cluster warheads, throughout the war).[xxv] Indiscriminate targeting of civilians not engaged in hostilities constitutes a war crime.[xxvi] It also perhaps constitutes a crass strategic error in this case – the oil refineries north of Baku are within range of Scud Bs fired from Nagorno-Karabakh and attacking them could have posited an existential threat to the Aliyevs’ rentier regime possibly inducing them to talk: nevertheless, there was just one report of an attempted attack, on 14 October.[xxvii]

The missile attacks proved a strategic distraction as the war was decided on the ground through some hard fighting. Combat seems to have been highly attritional, an advantage for Azerbaijan with its superior numbers, but for an outnumbered force supposedly being pulverised from the air by flying Terminators, the NKDF fought hard to the very end, inflicting heavy casualties on the Azerbaijanis in positional mountain combat in which, unsurprisingly, light infantry featured prominently and in which the Azerbaijanis clearly did not have it as easy as many Western pundits claim. Official figures for killed in action from September to November are 2,783 for Azerbaijan and 2,317-2,425 for Armenia; by way of comparison, Israel suffered 2,656 KIA in the Yom Kippur War of 1973.[xxviii] Azerbaijani troops – mainly conscripts and recently-mobilised reservists with varying quality of training – had to fight uphill through prepared lines and other positions, frequently under artillery fire from the defenders, to take a long series of fortified towns and villages, lists of names being broadcast nightly in the media and turned into memes on social media.[xxix] Typifying this was the battle for Fuzuli on 29 September, which began with Azerbaijani Special Forces occupying the hills around Fuzuli, cutting its communications with the intent of forcing the NKDF to abandon the town without serious fighting, a pattern repeated throughout the war in the s outh.[xxx] However, this was followed up by a conventional advance combining tanks with mechanised forces, the Azerbaijanis suffering some considerable losses of armour and people as the NKDF carried out a fighting retreat, most losses being to portable ATGMs or artillery, an indicator of why the offensive in the north bogged down into positional warfare for the rest of the war.[xxxi] Azerbaijani forces had little close air support throughout the war: Azeri Mi-24s were not committed until 5 October, fully a week after the offensive began and their contribution seems to have consisted of firing rockets from maximum range, aimed at saturating areas rather than hitting specific targets and much of their heavier artillery, the MRLS, for instance, seem to have been committed to barraging Stepanakert rather than supporting the advance in the south.[xxxii] NKDF forces were able to organise local counterattacks up to the closing days of the war, such as that on an Azerbaijani tank battalion which got within four miles of the Lachin corridor only to be driven back by MRLS fire with several tanks destroyed and the decisive moment of the war, the recapture of Shusha, actually took place in overcast weather restricting the use of UCAVS and any other air support, and was brought around by lightly-armed Azerbaijani Special Forces advancing through the mountains behind the city to take its garrison by surprise.[xxxiii]

So, what part did the UCAVs play in all this? Certainly, an important one but perhaps not as all-encompassing as some claim. ‘Drone’ strikes began on 27 September, concentrating heavily on the NKDF’s short-ranged air defences, over the first few days destroying fourteen SA-8s/SA-13s, four SA-10s and eight air defence radars, all struck by TB2s with a SA-3 near Stepanakert being taken out by a Harop. With these strikes we offer our first point of contention, that these were elderly Soviet-era systems largely incapable of tracking targets with radar signatures as small as the UCAVs presented; moreover, NKDF radars were jammed extensively by Turkish-supplied systems and, indeed, further TB2 attacks took out the NKDF’s two Russian-supplied Repellent 1 counter-UAV systems which detect incoming UAVs by their control signals.[xxxiv] Consequently, NKDF forces often operated blind and without any serious air defence, a major force multiplier for the Azerbaijanis and their Turkish allies. Following this, the UCAVS were switched to attacking ground targets with priority given to artillery, MRLS, tanks and supply dumps and vehicles moving along roads behind the battlefield.[xxxv] Noteworthy incidents included that of 1 October, where a NKDF armoured force massing for a counterattack in the north came under sustained UCAV attack, losing ‘many tanks’ and by 23 October independent open-sources were estimating that the NKDF had lost 144 T-72s, 35 BMPs, 310 soft-skinned vehicles and 116 artillery pieces; which system was hit by what was unclear, but the same sources estimated Azerbaijani Harops had destroyed 34 targets and ground-launched ATGMs 21 with much of the balance, presumably, going to the TB2s.[xxxvi]

Impressive figures, but context helps here. UCAV footage featured prominently on Azerbaijani television and on electronic billboards in Baku, forming a key part of Azerbaijan’s intensive multi-lingual social media campaign, ‘snuff movies’ posted to YouTube and Twitter selling the narrative of Azerbaijan smashing its detested foe with impunity. There is little surprising here for the experienced eye, the films reinforcing the eternal adage that on the modern battlefield, being seen is being hit, one tank, AFV or truck after another getting ‘plinked’ in masses of smoke and flame by MAM-Ls or other systems the Bayraktars and other UAVS are spotting for. Just as evident is the poor discipline and drill of the crews; on one level, there was the extensive use of mobile phones with GPS and even postings to social media by soldiers of both sides, showing their locations for all to see; on another, the films show target after target moving and sometimes parked in the open in broad daylight with no camouflage or overhead cover when stationary or even in prepared positions, so making life easier for the UCAV pilots than better trained and disciplined opposition might do.[xxxvii] Michael Kofman of the Wilson Centre and Jack Watling of the Royal United Services Institute (RUSI) both investigate these phenomena in detail, contending that NATO/Western mechanised forces would be just as vulnerable when manoeuvring, particularly given the modern battlefield is swept by long-range radars and electronic surveillance and most camouflage will not protect against modern infra-red detection systems. However, solutions differ, Kofman arguing for smaller forces with more extensive air defence, while Watling contends that ‘swarming’ can be forestalled by dispersing forces in non-tactical phases such as Armenia and Azerbaijan did not do: consequently, the tank might is not obsolete but needs to fit into a new tactical system emphasising moving dispersed but concentrating rapidly for the tactical phase.[xxxviii] To this we can add that NATO or Russian formations are likely to have air defences far more capable of engaging small, evasive targets: it is also remarkable that despite the Azerbaijani acquisition of the TB2s being ‘open book’, NKDF forces were not equipped with smoke generators or laser detection systems which would have given their vehicles at least some degree of forewarning and protection against laser guided weapons like the MAM-L.[xxxix] Claims of yet another ‘revolution in military affairs’ might therefore need some qualification.

Conclusions

The Nagorno-Karabakh war is significant for those with professional or academic interest in 21st-century war in that, alongside recent action in Ukraine, Iraq and Syria, it suggests claims about the death of conventional warfare might be premature. Rather, it might just have a present and a future, too. The war saw a series of conventional operations carried out by the Azerbaijan Army, with assistance from Turkey, attaining most of the geopolitical objectives their President and Commander in Chief set them: these operations hinged on high-intensity attritional combat aimed at seizing key ground, culminating in the securing of a major centre of gravity unhinging the entire Armenian position in Nagorno-Karabakh – so the basics of land warfare still count in scenarios like this one. Casualty figures were high for small countries engaged in a short war, but given we are dealing with two intensely nationalistic cultures fighting over territory, and the public and political reactions on both sides, the sacrifice seems to have been deemed worth it at least by Azerbaijan. However, when divorced from this they indicate eternal issues of attacking prepared positions in rough country and the need to coordinate infantry with support fires, something which might not have been done very efficiently here.

The much-hyped UCAVs contributed to Azerbaijan’s success in two ways, both important but hardly ‘revolutionary’ in that they provided a cheaper means of achieving things done traditionally by manned aircraft. First was suppressing enemy air defences in the opening stages, opening the Armenians up to the second impacting factor, deeper attacks destroying armour and disrupting lines of communication, tipping the balance of attrition in Azerbaijan’s favour in what was still a costly win for them. Here is something else transferable to other scenarios: ‘swarms’ of small UAVs and UCAVS might be a good – and cheap – investment but the conditions for best use need to be present, one being tactically inept opposition with air defences which can be overwhelmed early by swarming SEAD attacks; dealing with, for instance, the layered air defences forming an umbrella over large Russian formations might be a different problem. It might also be that the real messages here are not so much about the future of armour as that of tactical ‘fast air’ and attack helicopters, the Bayraktars in particular producing similar desired effects as these systems have for two generations for a fraction of the cost and with systems – unlike fast air and attack helicopters – their users could afford to lose and were cheap to replace. That might be the real transferable military message here while the political one might concern the acquisition policies of certain NATO countries, such as the UK, based on buying small numbers of highly expensive systems, and what those systems offer in reality.

My thanks to Miss Jamila Mammadova, Dr Mark Baillie, Brigadier Ben Barry, Major Sean Cronin-Nowakowski and others for their input to this paper.