- 1 The idea of the paper was first presented in a symposium entitled "The state and power in the Maghr (...)
1 Relationships between the state and tribal organizations have attracted the attention of numerous researchers. This paper discusses the development of a standing army in the al-Maghrib al-Aqṣā region, which roughly corresponds to the Kingdom of Morocco today, and analyzes the activities of standing cavalry troops called the Makhzaniyyah who played a significant role in the state taxation of the tribes. These troops existed as part of the armies of both Muslim and Christian rulers starting after the middle of the 15c. at the latest. Our study of these troops has special importance because it not only shows the development of state organizations in this early period but also indicates the plurality of the state for the tribespeople at that time.
- 2 According to Khoury and Kostiner, “because the term tribe has been used to describe many different (...)
- 3 We use a definition by Cory (2010: 454).
- 4 In reinterpreting the achievement of colonial historiography, which stressed the archaic nature of (...)
2 Regarding the historiography of the region, researchers have tended to consider the development of state organizations as a process in which tribes became subjugated by the central government on the basis of an assumption that the tribes were essentially non- or antistate, despite the diversity on both sides.2 From this perspective, they generally see the 16c. as a turning point. The period preceding this is that of the Berber dynasties, beginning with the Murābiṭūn (1061-1147), according to Laroui, a famous Moroccan historian, who stated that what the dynasty initiated in the 11c. had already disappeared by the time of the Waṭṭāsids (1472-1549, 1554) (Laroui, 2001: 221). This period is seen as a pre-state phase in which the most powerful tribal group of the region subdued or received approval from the other tribes and the cities. The subsequent period begins with the Sa`did (1510-1659) dynasty when Sharifian rule, or rule by sharīfs who claimed lineal descent from the Prophet Muḥammad,3 was established. The sharīfs of this branch, who lacked a tribal background, succeeded in mobilizing the tribespeople in the south of the country for Jihad with the support of Sufi saints, who had religious authority over society. The victory of the sharīfs over Portuguese invaders who had occupied the coastal regions enabled them to overthrow the Waṭṭāsid dynasty of Fez and reunify the country under their sole authority.4
3 In comparison to the Waṭṭāsid dynasty often condemned as the period of anarchy and stagnation, the period of the Sa`did dynasty is associated with innovative characters. This periodization was advanced by Cory, who also argued that the concept of a Moroccan state became widely accepted by most Moroccans from the 16c. to the 18c. (Cory, 2010: 453-61). In an attempt to synthesize the development of a Moroccan state organization, Shaqīr, a Moroccan political scientist, also highlighted the importance of the discontinuity between the Berber and the Sharifian dynasties and called the shift from the former to the latter in the 16c. a “political leap (al-ṭafrah al-siyāsiyyah).” The author provided several measures by which the new Sharifian state attempted to subjugate the tribal societies, such as the monopoly of the status of the monarch, the control of the religious discourse of the `Ulamās and Ṣūfīs, and the separation of military and administrative organizations from tribal ones (Shaqīr, 2006: 225-44). However, it is difficult to deny that a dynasty should prevent outsiders from approaching its throne. Moreover, it is well known that the Marīnid dynasty (early 13c.-1465)—the predecessors of the Waṭṭāsid dynasty—had already tried controlling the `Ulamās by introducing Madrasah (Shatzmiller, 1976). These two simple points strongly suggest that the Sa`did dynasty cannot be taken as the departure point of the incipient modern Moroccan state.
- 5 Although there are already many specialist articles about military organizations in this region, th (...)
4 Regarding the military organizations, discussion about the differences between the Berber and the Sharifian dynasties is more complicated and thus deserves closer scrutiny. Shaqīr’s scheme, which is indeed very clear, abstracts many changes that had occurred over several centuries. He argues that the state achieved independence from the tribal societies by introducing an army separate from the tribal societies and emphasizes the importance in the Sharifian dynasty of a standing army comprising foreign elements such as Turks and blacks (Shaqīr, 2006: 242-44). Dziubiński, in his study on the Sa`did army, even argued that before the dynasty was established, this region had no standing army (Dziubiński, 1972). However, upon closer inspection, these accounts have turned out to be unsubstantiated. In fact, Khaneboubi’s study of the Marīnid governmental systems demonstrates with numerous examples the presence of several military units comprising professional soldiers (Khaneboubi, 2008: 356-65). Thus, it is incorrect to stress the innovative character of the Sa`did army in this regard, and we need to investigate under what circumstances a standing army was introduced to the military system of the country.5
- 6 We discuss this qā’id later in detail.
5 In this paper, we integrate two lines of sources—Arabic and Portuguese—that have been studied separately. We first consider why the Makhzaniyyah came to be called by this name by examining Arabic documents from before the Waṭṭāsid period to identify indivisible associations of the units with state taxation. Then, we examine the activities of not only the units belonging to the armies of Muslim dynasties but also those that served the Portuguese king in the Dukkālah plain on the Atlantic coast. In this part, we rely mainly on Portuguese letters written from the late Marinid to the Sa`did period. In particular, accounts of the activities of the Makhzaniyyah, who fought under the direction of Yaḥyā Ū Ta`fūfut6 in the Waṭṭāsid period, are closely examined. Finally, we evaluate the activities and characteristics of the troops in relation to state control over tribal societies.
- 7 For the development of the makhzan, see Buret (1991) and al-Qablī (2005).
6 Etymologically, the term Makhzaniyyah derives from another, more well-known term, makhzan. This word originally meant a storehouse, and then it came to refer to historical North African government institutions. In Morocco, it is still used in reference to the `Alawī king and his inner circle. In the medieval Maghrib and Andalus, taxes levied by the states were stored in the makhzan. From the Murābiṭūn to the Muwaḥḥidūn dynasty (1130-1269), the word gradually came to denote the storehouses and treasury of the dynasty and was used in reference to financial affairs.7
7 In the Marīnid period, the term appeared more frequently in documents concerning tax collection and was associated with some words referring to taxes. For example, the makhzan’s waẓīf, or waẓīfah, meant a rent imposed on the lands of the makhzan called ribā’. A tax called the makhzan’s maghram was charged as a fine against participants in a revolt. It also meant a toll for merchants and an indirect tax for specified merchandise. According to al-Maqqarī, an author in the 17c., merchants who arrived at Fez during Abū `Inān’s reign in the mid-14c. paid a tax called the makhzan’s lāzimah (Khaneboubi, 2008: 286, 293, 294).
8 Moreover, the usage of the term occasionally emphasized a confrontation between the state and its subjects. For example, a loan fee on the real estate of a person who died without any inheritors was called a jazā’, which was classified as that of the makhzan and that of the subjects (Khaneboubi, 2008: 298). Ibn Ghāzī, a jurist of the Waṭṭāsid period, also contrasted the share (ḥaẓẓ) of the makhzan and that of the subjects in an anecdote about a tax collection contract in the Muwaḥḥidūn period (Ibn Ghāzī, 1988: 27). This term must have had a negative and oppressive connotation. A fatwā, or nonbinding legal opinion issued by a jurist of Fez in the middle of the 14c., reveals that there were people among the inhabitants of the city who had committed frauds in their gathering of taxes for the makhzan. In his response, the jurist checked whether “makhzaniyyūn (literally, those related to the makhzan)” indicated governors, guards, and soldiers who collected taxes unlawfully (al-Wansharīsī, 1981, 12: 63-66). Thus, the term supposedly suggested in the latter Marīnid period the state taxation entailing misconduct and injustices committed by various levels of people ranging from lower officials to governors.
- 8 Al-Qablī, 1987: 114; al-Sakkāk, 1892: 14. Although it is well known that slaves in Oriental history (...)
9 In the first years of the 15c., Ibn al-Sakkāk said in his treaty aimed at supporting the Prophet Muḥammad’s descendants that many people were facing injustices and humiliation from the police, servants of the governor, and slaves (wuṣfān) of the makhzan’s people.8 Although the text does not clarify the meaning of “makhzan’s people,” they should have been high officials of the Marīnid court. Additionally, João Álvares relates in his chronicle that when he traveled to Fez in 1437, people called the court the makhzan (mazagania) (João Álvares, 1959: 98). These testimonies suggest that the term had already been extended from the royal treasury of the dynasty to its court then. Probably because of the association of taxes levied by the dynasty with the qualifier makhzan, this term came to refer to the dynasty’s court, or to the government itself.
10 In that case, however, why did specific military units become called the Makhzaniyyah? The answer is that they received payment from the makhzan and played an important role in its tax collection. In the following sections, we consider the activities of the Makhzaniyyah, focusing on this point.
11 In Arabic sources written before the 15c., there is no mention of a standing army called the Makhzaniyyah. However, the term begins to appear in a few Portuguese texts relating to the military feats of Portuguese noblemen in northern al-Maghrib al-Aqṣā, where it means elite troops of the Marīnid and Waṭṭāsid dynasties. For example, we can find mention of these troops in the context of a siege of al-Qaṣr al-Ṣaghīr, a small port city bordering on the Mediterranean Sea. After the Portuguese occupied that city in October 1458, the Marīnid sultan `Abd al-Ḥaqq II besieged it. Gomez Eanes de Zurara, a contemporary Portuguese chronicler, reports on this battle: “the Moors, and especially those Marīnids and the Makhzaniyyah (mazaganis) surrounded the city (Zurara, 1978: 130).” On February 2, 1488, a Portuguese commander of Aṣīlah fought a battle with a Waṭṭāsid qā’id of al-Qaṣr al-Kabīr. Although this qā’id was accompanied by 550 Moorish cavalries in the field, he was defeated, and 50 of his Makhzaniyyah (mazaganys)—principal men—died (Pina, 1977: 945-46). These descriptions are so succinct that it is difficult to discuss the precise nature of these troops. However, the authors mention them by name, and Pina states that the Makhzaniyyah were the principal men of the qā’id’s army. Thus, we can guess that they were prominent troops comprising an elite cavalry.
12 Valentim Fernandes offers a more detailed account of this cavalry, saying in his topography written around 1507, the king of Fez deploys garrisons and soldiers to all the places that lie on the coast and in the vicinity of the Christians. The Moors call them the Makhzaniyyah (Machazanic). That is because makhzan (machazem) means a storehouse where all the rent of the king is gathered and paid. From this rent, the soldiers of each place receive a salary. For this reason, each place has its makhzan storehouse. The king does not take other things from the places mentioned above. All the same, the soldiers do not seize [from the places] anything because every day they receive the foods, and the horses are the king’s possessions. From these [places] the king would bring about 4,000 horses in addition to what he always brings with him. The king of Fez maintains 8,000 horses every day. All the horses are the king’s possessions. When he wants to assemble troops of Arabs and Berbers when necessary, he will assemble more than 40,000 horses, because there is a population consisting of Makhzaniyyah (machaniz), Arabs, and Berbers in a region from Aṣīlah to Fez (Magalhães Godinho, 2011, 1: 246).
13 This report demonstrates that the Portuguese understood makhzan as meaning a storehouse located in various places in Waṭṭāsid territory, or Bayt al-Māl, where all the tax income was brought initially. Thus, Valentim Fernandes thought that the name Makhzaniyyah came from the fact that they received from this makhzan all they needed for the defense of the frontiers.
14 However, we cannot determine whether the deployment of the Makhzaniyyah was limited to such frontiers where military confrontation with the Christian enemy was ongoing. Garrisons deployed in inland areas might have had the same name, but they would not have attracted a foreign author.
- 9 Castries, H. et al. (eds.). Les sources inédites de l’histoire du Maroc de 1530 à 1845, Portugal 2: (...)
15 Another mention of this unit appears in accounts of the third Waṭṭāsid expedition to the Dukkālah region in 1517. On July 29, Nuno Mascarenhas, the then commander of Safi, sent a letter to the king of Portugal, Manuel I, and reported that the Makhzaniyyah (marzaganys) of the Waṭṭāsid sultan had come to see Yaḥyā Ū Ta`fūft along with Sufi saints and a tribespeople of the region, accompanied by 20 riders, and had given him a message from the sultan.9 In a letter dated August 1, this commander wrote to the king that some tribal chiefs had returned from the camp of the sultan. Some Makhzaniyyah (marzaganys) had arrived with them to collect taxes from all the tribes (SIHM Portugal 2: 142-43). As the preceding expeditions in 1514 and 1517 had already devastated the region and made it difficult to maintain a large army, the sultan had to reduce his army’s size. Thus, he ordered the tribes that had come with him to go back and waited with his qā'ids and a few soldiers for some hostages from local Arab tribes to come to his camp near a lake (SIHM Portugal 2: 148). The Makhzaniyyah mentioned above should have come from this camp to the tent village of the Arab tribes. Hence, at least a part of the sultan’s cortege comprised the Makhzaniyyah, and probably the sultan trusted their fidelity and used them as elite troops.
- 10 The text of the manuscript here diverges from that of Ramusio edition that Épaulard’s translation f (...)
- 11 Leo Africanus regards the units of the Arab tribes more highly than the dynasty’s standing cavalry (...)
16 Leo Africanus also alludes to standing cavalry troops that protected the sultan, without using the term Makhzaniyyah. According to Leo Africanus, the king of Fez always had 6,000 cavalry soldiers who received a salary, 500 riding crossbowmen, and 500 riding shooters, and kept them prepared for his command. When the sultan went out to the countryside in peacetime, they took their places at a distance from him, and only 100 crossbowmen and 100 shooters attended him nearby.10 However, when he made war with Arab tribes and his enemies, his 6,000 cavalry soldiers were insufficient, and he needed help from Arab tribes that were obedient to him. These tribespeople participated in the war at their own expense and were numerous and much more experienced at fighting than the king’s cavalry.11
17 From this account, we can conclude that in the Waṭṭāsid army, standing troops protected the sultan and defended frontier regions against the Christian enemy. At least a part of them was called the Makhzaniyyah because they received their salary from the taxes paid to the treasury, or the makhzan.
- 12 Regarding the process of the Portuguese expansion in the region, the study of Būsharb remains the m (...)
18 Before examining details about the Makhzaniyyah in the Dukkālah region, we briefly summarize the history of Portuguese domination over the region.12 We can divide this expansion period into two phases. In the first phase, spanning the period between the expedition against Ceuta in 1415 and that against Graciosa in 1489, the Portuguese deployed armies against coastal cities in the northern al-Maghrib al-Aqṣā intermittently and occupied the lands around them.
19 Conversely, we can find a more peaceful mode of Portuguese advance into the southern part of the country. For example, local notables of Safi and Azammūr, two major coastal cities in the Dukkālah region, concluded peace treaties with Portuguese kings through emissaries sent to them. This approach seems to have enabled these cities to enjoy broad autonomy from the central government of the Waṭṭāsid dynasty without suffering direct and strict control by the Portuguese kings.
20 In the second phase, or the first decades of the 16c., the kingdom of Portugal under the reign of King Manuel I (r. 1495-1521) changed its strategy concerning this region. He sent many expedition armies to various places and succeeded in occupying Safi (1508) and Azammūr (1513), where Portuguese commanders and garrisons were stationed to control the cities and the neighborhood, thus realizing more direct rule over the region.
21 In the Dukkālah region then, the countryside population comprised Arab and Berber tribes, with the former being predominant. They inhabited tent villages called dawwār and lived a seminomadic existence following both the agriculture and the pasturage (Benhima, 2008: 139-43; 163-80). The Portuguese concluded a peace treaty with these tribespeople surrounding the occupied cities. Then, after establishing their rule, they began expanding their influence to more inland areas. In 1515, the Portuguese expedition army, comprising not only Christians but also Muslims, even attacked Marrakech, one of the most important cities in southern al-Maghrib al-Aqṣā (al-Qablī, 2011: 331-32). These conquests provoked counterattacks by the Waṭṭāsids, which carried out four expeditions in the region during 1514-1518. Such military campaigns targeted Arab tribes that had allied with the Portuguese, plundering these tribes’ lands and devastating local economies (Shinoda, 2014).
22 The organization of the Makhzaniyyah troops in service to the Portuguese kings came into force in this historical context and aimed at securing Portuguese rule over the conquered area and its populations. Būsharb, a specialist in the history of Lusitanian domination over the region in the 15-16c., stressed the importance of local elites, consisting of shaykhs and qā’ids. According to him, the Portuguese preserved the existing elites and used them to secure the loyalty of tribespeople in the countryside. The Portuguese king and his commanders followed the choice of a tribe’s people in the nomination of their shaykh, whose power was only formal and traditional. Conversely, only the king had the right of nomination of a qā’id and might choose him from outside the tribe. Such qā’ids exercised administrative power as representatives of the king and commanders in the countryside. They assumed the obligation of keeping the tribes in obedience, collecting and paying taxes, and facilitating Portuguese influence. They also had jurisdiction over the tribes through both Islamic and traditional law (Būsharb, 1984: 256-61).
- 13 For this qā’id, see Rosenberger (1993); Racine (2001); and Cruz (2002). Cosme (2004: 137-55) reveal (...)
23 Although this opposition between qā’id and shaykh seems too rigid to comprehend the reality as a whole, Yaḥyā Ū Ta`fūft should have been one of these qā’ids and probably the most typical.13 A Muslim of Berber origin living in the Dukkālah region, he came to the forefront amid civil strife among city notables over a government in Safi by participating in the assassination of its governor. In 1507, however, he left for Portugal after his defeat in this conflict. Later, in 1510, he returned to the city as the qā’id of Arab tribes of `Abdah and al-Gharbiyyah (Rosenberger, 1993: 22-30).
- 14 He indicates the source of this paragraph by only Būsharb (1984: 262), where we cannot find any men (...)
24 Some researchers, such as Būsharb (1984: 260) and Rosenberger (1993: 39), remark on the existence of a unit called by this name that served the king of Portugal under Yaḥyā’s command. However, they only briefly mention the unit’s name and number and do not comprehensively explore its activities. Conversely, Cornell, in his study of the Portuguese conquest of this region, makes a relatively long statement. He remarks on the Makhzaniyyah’s role in areas under Portuguese control, that is, the collection of taxes from the countryside, although his authority for this assertion is unclear.14 He does not, however, compare the Makhzaniyyah of Yaḥyā with those of the Waṭṭāsids and their predecessor, and evaluates the troops as “no more than a band of mercenaries” (Cornell, 1990: 386). In the chapters below, we show, using mainly contemporary letters, under what circumstances Yaḥyā’s Makhzaniyyah were organized and for what purpose this unit operated.
25 A Portuguese proposed the establishment of the Makhzaniyyah to the king of Portugal after the spring of 1514, and Yaḥyā implemented it. He had participated in a battle against a Waṭṭāsid expedition army led by al-Nāṣir, a brother of the sultan, that had occurred that spring near Būla`wān. On this battlefield, he commanded tribespeople from the Arab tribal confederations of al-Gharbiyyah and `Abdah. On May 19, Estevão Rodrigues Berrio, a commander of Portuguese battleships, praised the military exploits of Yaḥyā and suggested to the king the following:
26Your Highness must give him 100 cavalries for his guard so that Your Highness will be served, and the Moor will be estimated among the Arabs, and the things necessary for your service will be done because this Moor is Berber. For this reason, what he commands is not always executed.
27He then concluded that the king would profit greatly for just the small cost necessary to organize these 100 cavalries, and that Yaḥyā would be a great example for the Moors (SIHM Portugal 1: 555-56). However, this letter does not contain the term Makhzaniyyah.
28 This suggestion did not come to fruition, though, because Nuno Fernandes de Ataide, the commander of Safi, arrested Yaḥyā and deported him to Portugal on August 31. On September 12, he sent a letter to King Manuel I clarifying the reasons for his action. His explanation allows us to see that in August of that year, Yaḥyā had already begun organizing his guard corps, and that this corps had been called the Makhzaniyyah. Among the commander’s criticisms against Yaḥyā, he saw it as a problem that Yaḥyā bought many horses and formed many Makhzaniyyah (mazagamis), exceeding 100, and that he undertook this effort despite the commander’s objection. He added that Yaḥyā had intended to move without any authorization to a castle south of the Tānsīft River in the Shayāẓimah region, accompanied by all his troops and their families (SIHM Portugal 1: 623-33). On December 12, the commander again sent a letter accusing Yaḥyā of misdeeds. According to him, Yaḥyā had committed robbery against the tribespeople of the Dukkālah region with his Makhzaniyyah and deprived them of more than 2,000 ounces of silver. He even stated that with this money Yaḥyā had waged war against him, and thus against the king (SIHM Portugal 1: 662). These claims reveal strong displeasure on the part of the commander toward Yaḥyā that stemmed from his suspicion that the latter might rebel against the Portuguese and secede if he had cavalry troops.
29 In fact, the suspicions of the Portuguese nobleman, who repeatedly charged Yaḥyā with being a traitor and argued that he had conducted himself as king of the Muslims (SIHM Portugal 1: 379; 383), might have stemmed in part from the qualifier Makhzaniyyah, whose choice should have been made by Yaḥyā. In an ordinance he issued to a tribe in 1512, he named his treasury makhzan and called himself a sultan (SIHM Portugal 1: 318). As we have seen, the term already had implicated the state taxation and government in the 15c. This terminology could have given his fellow Muslims the impression that they were serving a legitimate Muslim sovereign and not a Christian king.
30 Although the commander stubbornly resisted the organization of the Makhzaniyyah, he did not mention in his letter anything about Estevão Rodrigues’s suggestion, despite its similarity with the force that Yaḥyā had organized in reality. This silence may indicate the king’s approval of its formation, and the commander may have abstained from openly expressing opposition to this plan because, without his approval, it should have been difficult for the qā’id to proceed with the troops’ organization against the commander’s will.
31 After Yaḥyā’s return to Safi from Portugal in July 1516, bearing an appointment letter from the king that stated his nomination as a qā’id, his guard cavalry troops began operating fully. His return was due to the critical situation that the Portuguese faced in the Dukkālah region following the death of Nuno Fernandes de Ataide that year and the massive defection of Arab tribes that followed. These two events meant the elimination of Yaḥyā’s enemy and an increase in his value as a Moorish leader capable of controlling the tribespeople.
32 This appointment letter is one of the most important documents concerning Portuguese domination in the region. In it, King Manuel I first appointed Yaḥyā as a qā’id of all the Dukkālah region and recognized him as having the same power and jurisdiction over the tribal confederations of `Abdah and al-Gharbiyyah as those he had possessed before his deportation. Then, he gave him the right to one-fifth of the booty acquired in a raid and juridical authority over the Muslims in compliance with the traditional customs known among the chiefs and peoples of the region. The king also provided him his standard and drum as a sign of Yaḥyā’s status as a qā’id of the entire Dukkālah region. After that, the king authorized him to receive 100 mounted lancers (lanças), who would always move with him for the king’s service. Each lancer was promised clothes and a feedbag (cevadeira) for his horse, which must have meant in reality fodder for the horse. This requirement to provide fodder was imposed on the Dukkālah region’s Moors as a part of their tribute, and it was delivered to the horsemen at the king’s granaries in Safi and Azammūr, or to other places that were more convenient for payment and satisfactory to Yaḥyā (SIHM Portugal 2: 7-8).
33 Thus, the cost necessary for these cavalry troops was prescribed to be paid from the taxes collected by the Portuguese king, as the cost needed for the Waṭṭāsid Makhzaniyyah was paid from taxes collected by the sultan. The king then placed on Yaḥyā an obligation to collect all the taxes levied on the population of the region and to transport these to the king’s granaries. Moreover, despite his decision not to accord Yaḥyā any competence to give any Moor an office or a charge, the king authorized him to appoint Moors for the collection of taxes. He stated, “in matters concerning the payment and collection of our tributes and taxes, he can delegate and order some Moors whom he considers competent to more rapidly and diligently make the payments mentioned above come, and in a time of necessity, make them executed (SIHM Portugal 2: 8-9).”
34 Thus, it appears that the king expected Yaḥyā to behave as his official representative for the inhabitants of the region, including the countryside, and to collect taxes among them. The mounted lancer troops were necessary for him to carry out his duties in the countryside where the Arab tribespeople lived. Because these seminomadic peoples residing in dawwārs continuously moved around the region and cultivated their lands, they could travel far away, accompanied by their families and livestock, and leave their lands and crops when their living conditions deteriorated due to war, for example. Troops provided with not only royal authority but also possessing mobility were necessary to control such people.
35 Like Estevão Rodrigues’ suggestion, this letter does not use the term Makhzaniyyah in mentioning the authorization to give Yaḥyā 100 lancers. However, in the documents we examine below, their authors, including Yaḥyā himself, always call the cavalry troops the Makhzaniyyah. Hence, we assume that Yaḥyā’s Makhzaniyyah means a unit comprising the 100 mounted lancers that the letter mentioned.
- 15 The Portuguese of this period call a camp composed of dawwārs of tribespeople azemel or azemell. Da (...)
36 The first mission Yaḥyā carried out after returning to Safi as the qā’id of the entire Dukkālah region was to force the seceded Arab tribes to conclude a peace with the king of Portugal again. In a letter Yaḥyā sent to the king on August 9, he reported that when he came back to Safi in July, among the Arab tribes of the region, `Abdah and Ecija had stayed in the azemel15 of the Sa`did Sharīf. All the tribes of al-Gharbiyyah had joined their azemel with that of Mūlay Nāṣir, a governor of Marrakech, and settled near the city’s ports. The leader of these tribes came to Safi in response to Yaḥyā’s demand on July 21 and entered a room where Yaḥyā; Nuno Mascarenhas, the new commander of the city after the death of Nuno Fernandes de Ataide; and other high officials were gathered. There a peace treaty was concluded as the king had commanded. A proclamation was written and announced in the city’s market in front of the chiefs and people of the tribes (SIHM Portugal 2: 24-25; Rosenberger, 1993: 42).
37 After the peace was negotiated, Yaḥyā continued to wage war against the enemy and to lead the tribespeople for around two years. In this period, he also led troops called the Makhzaniyyah and used them differently from other forces, e.g., on about March 20, 1517, Yaḥyā made a night raid against troops of the governor of Marrakech who had come to do battle with him. In this operation, he ordered his azemel to halt and advanced with only the cavalry of the Makhzaniyyah (mahazanya) (SIHM Portugal 2: 69). In the same year, from late March to early April, he fought with the troops of the Sa`did Sharīf. When the Arab troops began a skirmish, Yaḥyā advanced to another side of the battlefield with his standard and Makhzaniyyah (mahazanins) to intercept the Sharīf’s retreat (SIHM Portugal 2: 71). These incidents reveal the trust Yaḥyā had in them as skilled troops capable of operating in the middle of the night or the middle of a confused battle.
- 16 In spite of the censure the inhabitants of Safi made, the Portuguese did not abandon their human tr (...)
38 These troops may have had a nontribal composition based on their characteristics. In June 1517, Yaḥyā sent a letter to the king complaining about the treatment he had received from the Portuguese in Safi. At that time, information about the third Waṭṭāsid expedition had reached them, and they had suspected him of treason. In this tense situation, according to Yaḥyā, some Portuguese officials captured Moors, arguing that they might be Shāwīs. Yaḥyā claimed that if he did not have words of precaution and a letter from the king, they should make him a Shāwī (de mim fizeram Enxovio), as they did with one of his Makhzaniyyah (marzagani) (SIHM Portugal 2: 103-104). In the documents of this period, the name Enxovio meant an inhabitant of the Chaouïa region of today that lies to the north of the Dukkālah region. People of this area were not obedient to the king of Portugal and thus were available for the Portuguese to capture without breaking the peace that had been concluded with the inhabitants of the Dukkālah region. In another letter Yaḥyā sent to the king, he again protested against the Portuguese, who had taken prisoner one of his subordinates, saying that he was a Shāwī (SIHM Portugal 2: 108). It is hard to judge definitively whether the Portuguese allegation about the participation of a Shāwī in the Makhzaniyyah was true,16 but we may at least assume that the troops were considered to have been open to external elements to the tribal society of the region.
39 After the Waṭṭāsid expedition started in July, Yaḥyā’s situation deteriorated further, as the tribespeople of al-Gharbiyyah and `Abdah feared pillage and demanded that Yaḥyā give hostages to the Waṭṭāsids to stop the expedition. In response to his denial of their demand, they answered with a revolt against him. Conversely, the Makhzaniyyah seems to have shown fidelity to their qā’ids who were caught in a bind. On August 1, Yaḥyā sent to King Manuel I a letter written in such corrupt Arabic that it is difficult to understand the meaning. He reported that when he went to Sarnū, his home, the Arab tribespeople revolted against him, attacked makhzanī, and plundered the market and all the cereals he had stored in Sarnū. Then, they came to his azemel and tried to capture him. So, he got on a horse and escaped to Safi with makhzanī who wished to follow him (SIHM Portugal 2: 133-34). In September, Yaḥyā left the city and continued to wage war against the Sa`did Sharīf. According to Rodrigo de Noronha, a Portuguese from Safi who provided the last account about Yaḥyā, the number of his Makhzaniyyah had increased to 300 at that time, and he could not get the provisions necessary for his troops in the city (SIHM Portugal 2: 172). This report shows that a large number of his Makhzaniyyah did not stay in line with the Arab tribespeople in their revolt and instead followed Yaḥyā. Of course, there must have been members of the troops who abandoned him, as he himself indicated in his letter. However, their number increased somewhat after the revolt. That may be because the extent of defection was small, and tribespeople who did not participate in the revolt and followed him became counted among the Makhzaniyyah.
40 These accounts show us the nature of the Makhzaniyyah. They received a salary from the tax paid to the treasury of the king of Portugal and were directly under the control of Yaḥyā. Although this qā’id had commanded both the Makhzaniyyah and troops composed of Arab tribes, they had been organized separately and operated in different ways. Moreover, we can surmise that the Makhzaniyyah might have been formed outside the tribal organization of the region if we take into account the participation of Shāwīs in the troops.
41 Yaḥyā died in April of 1518 when he was attacked by `Abdah. However, the Makhzaniyyah troops continued to exist for a while. At the beginning of 1519, Nuno Mascarenhas sent a letter to King Manuel I in which he suggested the appointment of a Çayde, a nephew of Yaḥyā, as the commander of a fortress at Çoquyate, a location in the Ḥāḥa region, with his 20 Makhzaniyyah (marjaganys), 50 foot soldiers of Gornyz, and 20 Christian crossbowmen and shooters (SIHM Portugal 2: 223). This Çayde participated in a raid that Nuno Mascarenhas carried out on about November 9. Damião Gois stated that the “Çaide came with all his company or the Makhzaniyyah (marzagania) which they call them as such in their language (Góis, 10: 52).” So, we can guess that this chronicler likely understood these troops as comprising mercenaries.
42 Thus, we can assume that after Yaḥyā’s death, the command of the Makhzaniyyah passed to his nephew. Mention of the troops, however, disappears after that, and an account about the garrisons of Çoquyate indicates a reduction in the size of the troops as its numbers decreased in comparison to what Yaḥyā had maintained. This change seems to have had some relation to the Portuguese domination in this region. In 1517, Nuno Mascarenhas had already expressed his doubts about the Portuguese attempt to advance into inland areas, arguing the deteriorated situation (Rosenberger, 1993: 47). In the 1520s, a spate of commanders’ deaths and incidents involving their captivity transformed the organization of the garrisons of Safi and Azammūr, as they became composed of foot soldiers suitable for the defense of the cities, whereas they had mainly comprised cavalry before (Būsharb, 1984: 210-11). When the Portuguese plan to occupy inland areas from Safi to Marrakech became impossible, the importance of the Makhzaniyyah, which had been necessary to maintain the king’s control over the Arab tribespeople, was lost.
43 The existence of troops called the Makhzaniyyah was not limited to the Waṭṭāsid and Portuguese armies. On February 10, 1519, the commander of Safi reported that the governor of the Atlas Mountains, his two brothers, and a son of the governor of Marrakech came together with all their Arab tribespeople and Makhzaniyyah (todelos seus Alarves e marzaganys) to punish the Arab tribespeople who had continued to obey the king (SIHM Portugal 2: 227-28). This account suggests that the governor of each region of al-Maghrib al-Aqṣā had his own makhzan and Makhzaniyyah then.
44 An account of Mármol regarding the Sa`did army during the reign of al-Ghālib (r. 1557-1574) enables us to confirm this generalization about the Makhzaniyyah. The author explains the troops assigned to qā’ids of various regions as follows:
45All the other soldiers [than those stationed at Fez and Marrakech] are distributed to the qā’ids of the frontiers and their encomiendas. That is because the king always gives all his sons, brothers, and other principal persons who have kinship with powerful Berber tribes, or who are chiefs of Arab tribes, possession of lands and vassals in encomienda with which they could keep the cavalry, estimating for them many vassals for each squire (escudero) under whose charge the vassals are. The king has nothing more to do with them other than to order to give them authorization for the collection of rent and to send someone to call them so that they should come with their troops when necessary. The qā’ids benefit and collect their rent consisting of wheat, barley, oil, butter, meats of chicken, money, and other things that belong to them. After they store them in their storehouses (almazenes), they give [such things] to each one of their squires, who are called Makhzaniyyah (Magazanis), according to the quality of his person and the people he must sustain in his house (Mármol, 1599: 96v).
46Here, we can see continuity with the Makhzaniyyah in the Sa`did period. A qā’id of each place was obliged to collect various taxes and store them in his warehouse, and then give his Makhzaniyyah a salary in proportion to his quality and the number of his family members. Although a lack of contemporary Arabic documents makes it difficult to verify, since the late Marīnid period, a standing cavalry unit called the Makhzaniyyah was supposedly deployed in regions of al-Maghrib al-Aqṣā. The troops were used to defend frontier regions against attacks from infidels and to control tribespeople in the countryside. Then, this military system was followed by the Waṭṭāsid and Sa`did dynasties, thus emphasizing continuity in the military system of the area from the Middle Ages to the modern age.
47 The relation between the Portuguese and the tribespeople of the Dukkālah region was complex insofar as rivalries among tribal confederations in the region made it difficult for the former to keep this relation stable over the long term. It became more difficult due to Muslim counterattacks against the Portuguese occupation. In these circumstances, the Portuguese attempted to conclude peace treaties with the tribespeople in the countryside, collect taxes from them, and mobilize them for military actions against those unwilling to obey. This role was played by a standing cavalry unit called the Makhzaniyyah, whose members were Muslims. To accomplish the objectives mentioned above, it was necessary to have troops with experience in the field and with mobility. They received a salary from the treasuries of the king of Portugal and controlled the tribespeople as a coercive force of the king in the countryside of the region.
48 We can confirm the existence of troops called by this name starting in the late Marīnid era and going up to the middle of the Sa`did era. They were used not only by the sultan but also by local governors and commanders, including Yaḥyā Ū Ta`fūft. These troops had a common peculiarity in that they were always a standing cavalry and received a salary from taxes, thus demonstrating this military organization’s continuity despite two dynastic changes in the period mentioned above.
49 In this paper, we examined the relationship between the state and the tribes. Regarding this central question in the history of al-Maghrib al-Aqṣā, we examined a marginal group: Muslim cavalry troops who served a foreign sovereign (the king of Portugal) in a coastal region distant from the capital of the reigning dynasty (the Dukkālah region). This approach enabled us to bring to light the process of development of a standing army as a coercive force against the tribal society in a period when the power of central government had been undermined. Moreover, it suggests the nature of the state in that period. Although the idea for cavalry troops under a Muslim commander came from a suggestion made by a Portuguese, it was likely chosen by Yaḥyā from among existing military systems as the most appropriate one for the situation. Yaḥyā’s Makhzaniyyah could have worked for an assertion that he was the holder of power in the region because he had his—in reality, his king’s—makhzan or state organization, whereas he was nothing more than one of the leaders of tribal groups. This plurality of makhzan and Makhzaniyyah suggests the image of the state for the inhabitants of the country should have been plural, composite, and distant from the Moroccan national state.
50 The Makhzaniyyah went forth in advance of troops comprising Turks and renegades in the Sa`did army, and famous Bukhārī troops consisted of blacks in the `Alawī army, as it was a standing unit organized separately from troops of local tribespeople in order to control the latter. Thus, we can regard its creation as part of the process of the state’s separation from the tribal society in al-Maghrib al-Aqṣā. However, the size of the troops remained small at this time, and even the Waṭṭāsid army relied upon an allied tribal organization for a large-scale expedition. Therefore, we may conclude that the creation of the Makhzaniyyah did not have sufficient impact to change the existing military system of the country, but that it demonstrates that the change that occurred in the military system from the medieval to the modern period was a continuous one rather than a rupture, as Shaqīr argues.