The Qapu Khalqi (‘Gate-people’ or ‘Court-people’), also known as the Qapu Kullari (‘Gate-slaves’ or ‘Court-slaves’) was in origin the sultan’s own army of dependent household slaves, ‘originally manned by prisoners and mercenaries, Moslem and non-Moslem alike’ as Stanford Shaw puts it. Its alternative name gave rise to the word Capiculari which often occurs in western sources, and the ‘Qapu’ element of both forms was the origin of the word ‘Porte’ which was soon being applied to the central Ottoman government and the sultan’s standing army alike by Byzantine and European chroniclers. Though it later came to include the Janissaries, the Qapu Khalqi was initially always mentioned separately from them, and it is under this name that the sultan’s slaves are recorded fighting at Ankara in 1402 and Varna in 1444. Doukas too, when recording the presence of the Porte at the Battle of Nicopolis in 1396, probably intends the Qapu Khalqi (or perhaps the Qapu Khalqi and the Janissaries, since any distinction between the two may not have been clear to him). He describes them on this occasion as ‘the Turks called porta, that is palace guards, who were redeemed slaves from diverse Christian nations and numbered more than 10,000’.
The principal military element of the Qapu Khalqi was initially comprised of several companies of cavalry, collectively known as the Qapukulu Suvarileri. There were in time 6 of these, the first of which were established from his own salaried horsemen by Kara Timurtash, beylerbey of Rumelia, whom we have already met. This was probably in 1376. These were known as the Ulufeciyan or ‘Salaried men’ (‘mercenaries, that is to say,’ as the ‘Ordo Portae’ succinctly observes); they were divided into left and right — i. e. Ulufeciyan-i yesar and Ulufeciyan-i yemin — indicating their battlefield position in relation to the sultan, thus constituting two companies. The next two units, established soon afterwards, were the Gureba, similarly comprised of Gureba-iyesar and Gureba-iyemin but this time made up of Moslem mercenaries from further afield, mostly Arabs, Persians and Kurds, but seemingly occasionally Tartars (see below); hence their name, meaning ‘Strangers’ or ‘Foreigners’. The Ulufeciyan and Gureba together comprised the Bolukat-i Erba’a, meaning ‘Four Boluks’ (or ‘Four Companies’). The final two companies, the Silihdars (‘Weapon-bearers’) and the Sipahi-oghlan (‘Sipahi children’) were probably established in Mehmed Ps reign, thereafter constituting the elite of the corps. They drew up on the immediate right of the sultan in battle, and along with some of the Ulufeciyan were recruited from among the Janissaries and household slaves (these both being drawn from the adjami-oghlan, or ‘foreign youths’, raised by the devshirme system mentioned above; the Sipahi-oghlan appear to have been drawn specifically from among slaves who were the children of noble families). Each of the six companies of household cavalry was commanded by an Aga, and in the 16th century at least they were organised in squadrons of 20 men. The ‘Ordo Portae’ tells us that the Sipahi-oghlan were paid 20-25 aspers per day, the Silihdars 10-20, the Gureba 6-10, and the Ulufeciyan 5-6. They were paid quarterly like the Janissaries.
An early reference to the strength of the Ulufeciyan and Gureba is to be found in the chronicle of Johann Schiltberger, an eye-witness on the Ottoman side at the Battle of Ankara, who describes how at the end the sultan stood with a single body of 1,000 cavalry — doubtless the Four Boluks. However, according to a 16th century source the Four Boluks originally numbered 2,400 men, though this may be a mistake for all 6 companies. Certainly the ‘Ordo Portae’ records the strength of the latter in Mehmed IPs reign as 700 Ulufeciyan, 400 Gureba, 600 Silihdars and 600 Sipahi-oghlan, adding up to a total of 2,300, while Bertrandon de la Brocquiere (who refers to the Ottoman household troops — probably excluding the Janissaries — as comprising 5,000 salaried cavalry and infantry in all) mentions in another passage that Sultan Murad II had ‘2,000 or 3,000 slaves of his own’, among whom there were many Christians — doubtless a reference to the household cavalry. Janus Lascaris too, later in the century, records that there were at the Porte 3,000 ‘knights who call themselves slaves’ who are ‘limited to a maximum strength of 5,000’. Arnold von Harff, who visited Constantinople in 1499, gives figures of 600 and 700 respectively for the Silihdars (‘mostly renegade Christians’) and the Sipahi-oghlan (‘who were his bodyguards’), thereby basically confirming the ‘Ordo Portae’, adding that there were also 7-800 trainee Silihdars, presumably adjami-oghlan recruits. Chalkokondyles, on the other hand, mentions a unit of only 200 bodyguards who are probably the Sipahi-oghlan, or an element thereof, and 300 others (later 500) who are probably the Silihdars. Mouradja d’Ohsson cites figures for the latter two companies under Mehmed II that are ridiculously high (10,000 and 8,000) even if one allows for the armed and mounted slaves that 16th century sources relate were maintained by household cavalrymen (5 or 6 by Sipahi-oghlan, 4 or 5 by Silihdars, and 2 or 3 by Ulufeciyan).
In addition to these companies the sultan also had a personal cavalry guard, the Micteferrika, which derived its name (meaning ‘Separated’) from the fact that it was chiefly comprised of the sons of vassal princes, who were effectively held as hostages but received training whilst in the sultan’s service. They performed military service only whilst on campaign, and probably numbered about 100 during this period. There were also special infantry guards called the Solak (‘Left-handed’) and the Peyk (‘Messengers’), both recruited from amongst the Janissaries (the former were part of the Cema’at) and both commanded by bashis. The Solak marched on the sultan’s left and right, and their name derives from the fact that those on his left were actually left-handed; it is they who Arnold von Harff is referring to when he talks of ‘foot-soldiers, like archers in France, who always follow him or run after him, and in towns they keep order on the left hand.’ The dozen archers Brocqui&re saw returning from a hunting trip with Sultan Murad II were probably part of the Solak (the 50 horsemen recorded on this occasion probably being Miiteferrika), while the 20 or 30 ‘baston’-armed
Slaves that he reports elsewhere as gate-guardians were possibly Peyks (but more probably Kapicis or Cavus — see below), since this unit appears to have been armed with axes or maces. The ‘Ordo Portae’ gives the Solak a strength of 80 men.
Mihailovic provides one of the most comprehensive breakdowns of the Qapu Khalqi military establishment that we have for this period, fuller even than that of the ‘Ordo Portae’, presumably describing it as it stood late in Mehmed IPs reign, though his figures do not always even come close to tallying with those quoted above. He describes the household cavalry as comprising Gureba of 600 Tartars (presumably Grim Tartars) in two boluks of 300 men each; Ulufeciyan, of identical size to the Gureba; Silihdar of 300 men (i. e., only half the size quoted by the ‘Ordo Portae’), whose task ‘is to lead a horse before the sultan when necessary’; and Sipahi-oghlan, also 300-strong. The last he describes as ‘the highest’ of the household regiments: ‘They have the task of carrying swords after the sultan, also having girded themselves with bows and arrows whenever necessary. It will fall to one of them to bear [a sword] once every year or two. All of them are mounted.’ The Solak, whose task he says is ‘to walk before the sultan with bows’, he puts at 60-strong. In addition he lists another regiment, the Kapici, comprised of 200 gate-guardians under two kaptci-bashis; the ‘Ordo Portae’ also mentions these ‘amongst the Janissaries’, though it puts their strength at only 80 men, paid 6-8 aspers a day. Since it is known from other sources that they performed the function of messengers as well as gatekeepers it seems likely that they were, or at least had once been, identical with the Peyk mentioned above.
Of the household regiments in general, including the Janissaries, Mihailovic reports that ‘they have their task in common: at night to lie near the sultan, and they take the night watch silently. Whether there is rain or snow, winter or a blizzard or whatever sort of weather, each must remain in his place, each night 50 and when necessary sometimes 100. And none need concern himself about any weapon; the sultan provides whatever befits each of them according to his rank, whether cavalryman or foot-soldier — armour for horses and assorted weapons, according to their custom.’ In addition he lists further elements of the Qapu Khalqi that have not yet been mentioned, comprising the Casni-gir or ‘Tasters’ (80 men); Icoghlan, basically chamberlains (50 men); Mirahur or grooms (200 men, responsible for the 2,000 horses which the sultan distributed ‘as necessary, with saddles and accoutrements, especially when there is a great battle’); Cebeci or ‘Armourers’ (60 men, founded by Mehmed II and responsible not only for making the armour and weapons used by the household troops, but charged with their transport too, even serving in the field as handgunners); Mehteri or ‘Tent-pitchers’ (60 men, not to be confused with the Mehteran-i tabl ii alem, ‘The Military Band’, for which see page 106); and lastly cooks, camel-keepers and musicians, numbering 200 men in all. Finally, there was one further body which Mihailovic does not mention within his chapter on the Porte but does refer to elsewhere; this was the Cavuslar, described as comprised of ‘courtiers on armoured horses’, who in battle performed the duties of sergeants-at-arms, being sent out by the sultan ‘to observe who is doing any brave deed and how the battle is going. And each of them holds a mace in his hand, urging [the troops] into battle. . . And wherever they are, it is as if the sultan himself were there; and everyone fears them, for whoever they praise in front of the sultan will get advancement, but woe betide anyone they criticise. Their leader is called the Cavus-bashi.' It is thus the Cavuslar to which Pseudo-Sphrantzes is referring in his version of the final assault on Constantinople, where he describes how when the Turks fell back ‘the sultan’s military police and court officials beat them back with iron maces and whips’. The ‘Ordo Portae’ puts their strength at 80 men.
Other elements of the Qapu Khalqi not listed by Mihailovic were the Topcu ocaki (‘Ocak of the gunners’) and Top 'arabaci (‘Gun-carriage drivers’) — for both of which see the section on artillery — plus the Humbaraciyan (‘Mortar-men’) and Lagimciyan (‘Sappers’). The latter two, both established by Murad II (but see page 14), were organised identically: they were each divided into two elements, one of which received a regular salary and was attached to the Cebeci and was thus part of the Porte proper, while the other was scattered throughout the provinces, supported by timars and effectively coming under the command of the provincial governors. In addition the Qapu Khalqi element of the Mortar-men was further subdivided, with those who actually operated the guns on campaign being attached to the Topcu ocaki. Each of these various companies was organised into ortas.
The irregulars: ’azabs, akinjis and others
The ’azabs (their name effectively means ‘bachelors’) seem to have originally evolved as marines in the amirates of Aydin or Menteshe in the late-13th or early-14th century, but in the Ottoman era the term generally applied to the masses of foot-soldiers recruited in whatever numbers were required for the duration of a campaign. According to Ottoman sources there were as many as 20,000 ’azabs at Ankara in 1402 and Constantinople in 1453, while Mihailovic records that Anatolia and Rumelia could field 20,000 each. In
Battle they were positioned in front of the Porte and feudal sipahis. They were mainly Turks and seem to have been provided and equipped on much the same basis as the yaya (one source even mentions Anatolian yaya and Rumelian ’azabs fighting side by side at Kossovo in 1389, so what differences there were must have been insignificant). In Bayezid IPs reign they were raised on the basis of each 20 khanes (households) providing one man, who was supported during the course of the campaign by money and provisions supplied in lieu of taxes by those who stayed at home. Though mainly bow-armed, their equipment was far from uniform; Brocquiere describes those of Rumelia as ‘some having swords without bows, others without swords, bows or any arms whatever, many having only bastons [staves]. It is the same with the infantry supplied by Turkey, one half armed only with bastons. This Turkish infantry is nevertheless more esteemed than the Greek, and considered as better soldiers.’ (See also notes to figures 3 and 4.) By the mid - or late-14th century many were beginning to appear in addition as parts of the garrisons of Ottoman fortresses.
The akinji were a light cavalry force which derived its name from the Turkish word akin, meaning a raid or an incursion into enemy territory — hence the term akinji, meaning ‘raiders’, which first appears in the chronicles in 1362. They were Turcoman ghazis in origin, who in the course of time became established under virtual dynasties of local chieftains (uc beyi, or ‘frontier chiefs’) at strategic frontier posts throughout the Balkans, notably in Hungary, Serbia, Moldavia, Wallachia and Greece. They were conscripted (voluntarily or otherwise) as and when required, in exchange for the usual tax-exemptions, their leaders receiving hereditary timars. It appears from Mihailovic’s ‘Memoirs’ that most actually depended on horse-breeding for their livelihood; he says the Turks sometimes called them gogmary, apparently a corruption of Serbian konjari (‘grooms’). He adds that ‘if any of them does not want to go on a foray himself, he will lend his horses to others for half [of the booty]’. Another good description of the akinjis is provided by Mehmed II’s treasurer, the Italian Giovan Maria Angiolello, who accompanied the sultan on his campaign against Uzun Hasan of the White Sheep Turks in 1473. He says that the akinji {‘aganzi') ‘are not paid, except by the booty which they may gain in guerilla warfare. These men do not encamp with the rest of the army, but go traversing, pillaging and wasting the enemy’s country on every side, and yet keep up a great and excellent discipline among themselves, both in the division of the plunder and in the execution of all their enterprises. In this part of the army there were 30,000 men [Angiolello’s favourite figure], remarkably well-mounted’. Mihailovic provides us with the additional information that each akinji served with two good horses (‘he leads one and rides the other’), and other sources tell us that their arms comprised lance, bow and sabre. Organisation was decimal, under officers called tovijeler, who received fiefs. Overall command of an akinji raid was in the hands of the sancak bey from whose province the akinjis were raised and the foray launched, who was called the akinji bey for the occasion.
Other cavalry included the djanbazan (‘daredevils’), created either by Orkhan or in 1441 by Murad II to counter Janos Hunyadi’s first Balkan expedition. They usually served in the vanguard of the army and therefore were probably drawn from the ranks of the akinji, but some authorities claim that they may have formed the bodyguards of sancak beys and beylerbeys. A seemingly similar body of light cavalry called delis (‘madmen’) was raised in the late-15th century from amongst converted Serbs, Bosnians and Croats; for further details of these colourful soldiers see Armies of the Sixteenth Century. Other irregular cavalry were provided by yuruks (‘nomads’, still called tiirkmen in Anatolia, where most were to be found) and Crim Tartars, the latter being supplied initially as independent auxiliaries but later, under Mehmed II, as a result of a vassal relationship that was hammered home by force in 1475. As an indication of their potential, various sources claim that there were as many as 30-70,000 Crim Tartars in the Ottoman army that marched on Moldavia in 1484.
Additional provincial troops were provided by the derbentci, formally established in the mid-15th century, and the goniilluyan (volunteers), in both cases comprised of mixed cavalry and infantry recruited from the local population, to guard fortified road-houses (derbents) and fortresses respectively. The latter group were a mixture of Moslems and converted Christians, but the former were chiefly comprised of non-Moslems, though they included in addition many yuruks. An alternative name by which the derbentci became known in some parts of the Balkans in the 16th century was martolos (from the Byzantine armatoliki), but in this period the term indicated nomadic Vlachs of the Pindus mountains who received tax-reductions from the Ottomans in exchange for their military service as garrison troops, watching over the passes under hereditary kipitanoi. They are first mentioned under Mehmed II, but evidence indicates that they date back to Murad II’s reign. Mihailovic says some had armour and that their arms comprised ‘swords, shields, lances and also guns’, but that there were not many of them — only ‘several hundred’, martolos and voynuks combined.
World History
