Oikoi stratiōtikoi. Open Questions on Land and Military Service in Byzantium (c. 7th–10th centuries)

1 From the tenth to the eighth century: a retrospective

Chapter 49 of the De Cerimoniis’ Book II is a layered text composed of at least three sections.^[4] The first is a list of the salaries (rhogai) and contributions due to a number of dignitaries in the age of Emperor Leo VI (886 – 912)^[5]; the second is a very short list of payments pertaining to chartularioi and notarioi of the Constantinopolitan offices, which also refers to the age of Leo VI;^[6] the third gives instructions on how to settle Muslim prisoners who have converted to Christianity in the territory of the empire (περὶ τῶν αἰχμαλώτων Σαρακηνῶν τῶν ἐπὶ θέματι βαπτιζομένων).^[7] If in this last section, as has been suggested, a reference to the settlement of the Banū Ḥabīb in the theme of Melitene and the districts of Charpezicium, Arsamosata, Chozanum and Derzene after 936 is to be seen,^[8] then it is more or less contemporary with the age of Constantine VII. This third section, the most difficult to analyse, is in turn composed of three subsections. In the first one, the prōtonotarios of the thema in which the group of immigrants – made up of about 12,000 individuals including fighters, women, and children – had settled is informed of the amount of pay and food rations due to each individual soldier who was part of the group.^[9] The second deals with the possibility of one of the new immigrant soldiers joining an oikos, be it stratiōtikos or politikos, by marriage.^[10] Finally, the third subsection deals with the case where one of the Muslim warriors receives a plot of land himself in order to settle down (γῆν εἰς κατασκήνωσιν).^[11]

Starting with an influential essay by Hélèn Ahrweiler more than half a century ago, the idea has gained ground in historiography that the binomial oikos stratiōtikos / oikos politikos indicates a type of family on which the military burden was carried out, as opposed to a second type free of this obligation.^[12] Constantine VII does not specify whether the marriage between a Muslim warrior and a woman belonging to a ‘civil family’ changes the latter’s functional status, transforming it into a ‘military family’; he does specify, however, that both, by welcoming an immigrant into their household, would be exempt for three years from the synōnē (in the 10th century, a tax in kind paid by farmers) and kapnikon (hearth-tax not mentioned before Nicephorus I).^[13]

When did the distinction between oikos stratiōtikos and oikos politikos arise and what were the conditions that made it possible? Chapter 50 of Book II of the De cerimoniis contains a very useful piece of information in this regard. It specifies the office holders who, if a supply of recruits for a military campaign was called for, were exempt from the charge.^[14] Arriving at this point, Constantine VII makes an important clarification which is worth reading in full: “it should be noted that if among all the mentioned individuals who are exempt from the financing of the military campaign, one had an old military charge (παλαιὰ στρατεία) as long as the person linked to one of the above-mentioned offices is alive, his oikos is exempt. But at the death of the latter, it is necessary that the military charge be reassigned to his oikos according to the palaios typos of the office of military affairs” (ἰστέον, ὅτι ἐκ τούτων τῶν προειρημένων μὴ στρατεύεθαι, εἶχέν τις παλαιὰν στρατείαν, μέχρις ἂν ζῇ τὸ καταταγὲν πρόσωπον, οἱονδήποτε τῶν προειρημένων ὀφφικίων, ὀφείλει διαμένειν ἐξσκουσευόμενος ὁ τοιοῦτος οἶκος. τελευτήσαντος δὲ τοῦ τροσώπου, ὀφέλει ὀρθοῦσθαι ἡ στρατεία εἰς τὸν οἶκον αὐτοῦ κατὰ τὸν παλαιὸν τύπον τοῦ στρατιωτικοῦ λογοθεσίου). The translation of the expression palaios typos is by no means straightforward and has, in fact, been interpreted differently by scholars. Reiske translates typos with the Latin formula (“secundum veteram formulam rationarii militaris”).^[15] Ann Moffatt and Maxeme Tall understand the term as ‘ordinance’ (“in accordance with the old ordinance of the bureau of the logothete of the stratiotikon”).^[16] Gilbert Dagron, on the other hand, as ‘ancien tarif’ (“second l’ancien tarif de la caisse du stratiotikon”).^[17] This latter translation, which is philologically audacious, is justified based on the circumstance that the presence of clerics from the imperial palace in the list would show that it was a tax or contribution in kind, not a personal service, since clerics could not have carried arms. Such a justification, however, seems incongruous with the wording of the text, which expressly clarifies that the burden is not on the clerics, but on their oikos, and therefore could have been fulfilled by their relatives. It seems to me, therefore, that the learned emperor with the expression palaios typos alludes here to the existence of the logothesion tou stratiōtikou – in excessively modern language we could call it the office or accounting office for military affairs, based in Constantinople – of an ancient register (typos: matrix), containing the enlistment of all oikoi who exercised the strateia.

At least in the 9th and 10th centuries we know that soldiers who served in the provincial armies were listed both in registers that counted the conscripts for each thema, and in a single general matriculation (kōdix) preserved in Constantinople, which perhaps represented the original state of constitution of the military ranks. Both must have already existed in the first half of the 9th century, as some lives of saints show, such as the bios of St. Euthymius the Younger or that of Luke the Stylite.^[18] The former was born in 823/842 in the village of Opso, in Galatia, son of a stratiōtēs, his father στρατείᾳ καταλεγόμενος (“enlisted in military service”).^[19] His parent died when the saint was only seven years old, leaving the mother uncertain whether to burden her child with the obligation; in the end she made the decision, and Euthymius was enrolled in the stratiōtikoi kōdikes. Despite this, at the age of eighteen (in September 841), he nevertheless embraced the consecrated life on Mount Olympus. The family of Luke the Stylite lived in Phrygia in the 9th (or 10th) century. His parents were well-to-do peasants, but also obliged to military service (στρατείᾳ κουστωδίᾳ καταλεγόμενοι).^[20] Around the age of eighteen he was considered capable by his parents to take on the obligation (τὴν τῆς στρατείας ἐξυπηρειτεῖν ἐπήρειαν προεστήσαντο);^[21] the bios also takes care to specify that Luke was not maintained by the rations provided by the imperial administration (ὀψώνιον ἤτοι βασιλικὸν σιτηρέσιον), but by the revenues of his father’s estates (πατρικὸς οἶκος). Luke also embraced the consecrated life and was later ordained as a priest at the age of 24.^[22] Although in such a condition, the biographer specifies that he continued to exercise the strateia (ἐξυπηρετούμενος τῇ στρατιωτικῇ ἐπηρείᾳ).^[23] Evidently the service must have been connected to his family by inheritance, which implied that Luke had to entrust it to a relative who was not of ecclesiastical status.

Another hagiography, the Life of Saint Pancras of Taormina, is also very interesting for the purposes of this paper. The date of its composition is uncertain. The editor of the text, Cynthia J. Stallman-Pacitti, based on several pieces of evidence, believes it to have been written before 732/733 (or, at the latest, before 757).^[24] Nevertheless, the iconodule florilegium of the Parisinus Graecus 1115, compiled according to Alexakis in the 770s, includes an excerpt partially based on the Life of St. Pancras^[25]. While Stallman-Pacitti and Lamparidi do not see any clear indications in the Vita that it was composed during Iconoclasm, it can be observed that the author’s sensitivity towards church decoration and icons (chs. 3 – 5, 6 – 9, 18 – 23, 24 – 25, 60, 262 – 263, 309 – 311), together with his denunciation of false idols (chs. 30, 99 – 100, 108 – 109, 114 – 115, 134, 262 – 263, 324 – 325), do not completely rule out this possibility. Although in general the Life of Saint Pancras is a hagiography where imagination, as has been written,^[26] prevails over reality, it nevertheless contains some interesting elements from a social-historical perspective. One of them concerns the gathering of the army described in the text. One of the many fictitious characters in the Bios, a ship captain named Lykaonides (who after his conversion to Christianity took the name Epiphanios) goes to meet the governor of Taormina, Boniphatios. The latter, impressed by his story, asks to make the acquaintance with the man of God, Pancras.^[27] Epiphanios then suggests he convene an adnoumion in a plain near Taormina, whereupon Boniphatios proclaims it by sending heralds in the town and messengers in the province to inform all the men under his command (ὑποτασσόμενοι) of his decision.^[28] The troops, when assembled, receive from Boniphatios rations of foodstuffs and donations for their benefit (ὁ δὲ Βονιφάτιος τοῖς στρατιώταις σιτάρχισιν δεδωκὼς καὶ δονατίβας πρὸς ἀπόλαυσιν τῆς χρείας); the chilliarchs, centurions and pentecontarchi of the army are invited to a banquet.^[29] So Boniphatios, after reviewing the armed men, sends them back to where they came from, while he himself, together with 500 selected soldiers who made up his bodyguard, goes to meet Pancras.^[30]

The fictional and imaginative narrative of the Life prevents a clear recognition of administrative institutions and practices reflected in the story. The figure of Boniphatios, who is referred to at several points as ἡγεμών, ὑπατικός and τοποκτίστης, resembles that of a stratēgos, even though he resides in Taormina, and not in Syracuse.^[31] An alternative hypothesis is that the hagiographer was inspired by the model of a topotērētēs; several seals from this officer have been found in Sicily and Calabria, mostly dated to the eighth century.^[32] An administrative procedure, however, well recognizable in the Vita, is that of the adnoumion (in the text τὸ ἀδνοῦμιν). This word is not attested as a technical term in the military language of late antiquity and finds one of its oldest mentions in the hagiography of St. Pancras, together with the better-known Life of St. Philaretos the Merciful (written in the early ninth century but set in the previous century).^[33] The term seems to derive from the Latin locution ad nomen.^[34] According to the Suda lexicon, it would indicate the inscription of names in a register or census (ἀδνούμιον· ἀπογραφὴ ὀνομάτων παρὰ Ῥωμαίοις),^[35] while a gloss to the Basilici defines adnoumion as ‘to review by name’ (ἀδνούμιος, κατ’ ὄνομα διέρχεσθαι).^[36] In late republican Latin, nomen, among its various meanings, also qualified a debtor recorded in a book of accounts. (Cic. de off. 3, 59). From this arguably developed the meaning of adnoumion in a military sense which we find attested for the first time in the collection of laws of Athanasius of Emesa (second half of the sixth century), where it means ‘to review’ the military units (ἀδνούμια ποιεῖ τῶν στρατιωτικῶν ἀριθμῶν).^[37] In fact, just as in the language of credit, the registration of a person was done because he owed a sum of money, registration in the military rolls was done because one owed a service. Two Egyptian papyri from the early eighth century uses the verb ‘ἀδνουμέω’ with the meaning of ‘to supervise and inspect workers’; moreover, one of them uses the form ‘τὸ ἀδνοῦμεν’ like that of ‘τὸ ἀδνοῦμιν’ in the Life of St. Pancras (with exchange between ε and ι for homophony) is attested.^[38]

The practice of ‘cataloguing’, in the literal sense (i. e. ‘listing’), the families that were engaged in military service is perhaps reflected in the first mentions of the disputed term thema, themata (plur.) found in Theophanes’ Chronographia. Its final redaction, regardless of the thorny problem of authorship, stops in 812/813.^[39] At the points in the work where the term is used, it seems to mean both provincial areas where troops were stationed and the troops themselves.^[40] Although there have been those who have derived the word θέμα from an Altaic root tümän (‘a thousand men’),^[41] the opinion of Franz Dölger and Nikolaos Oikonomides seems to me preferable, who, on the basis of an explicit testimony of Constantine Porphyrogenitus,^[42] thought that the term originally meant katalogos, i. e. list of soldiers, from τίθημι (> θέσις), which among its meanings also means ‘to record’.^[43] But even if one disregards Theophanes’ testimony, the existence of a register at the military accounting office in Constantinople listing all oikoi required to perform military service can be inferred from a letter of Theodore Studite to Empress Irene in 801, which has already been adequately highlighted by historiography.^[44] In it, the energetic defender of images praises the empress for having granted the remission of a certain number of taxes. Among the beneficiaries of the measure are also – I quote – “the wives of soldiers who, already suffering from the loss of their husbands, will not have to mourn further for a miserable and inhuman imposition resulting from their death” (αἱ στρατιώτιδες τὸ οἰκεῖον πένθος ἔχουσαι τῆς ἀνδρικῆς ἀποβολῆς, οὐκ ἐπιθρηνήσουσιν τὴν ὑπὲρ τοῦ θανόντος ἐλεεινὴν καὶ ἀπάνθρωπον ἐξαπαίτησιν).^[45] In this case too we are faced with a witness that seems to indicate that there were special military oikoi to the point that at the death of the male owner the service passed under the responsibility of the wife.

If the evidence analysed so far is to be believed, the use of registers in which ‘households’ (oikoi) performing military service were inscribed should date back at least to the second half of the eighth century. This custom developed against the backdrop of an acceleration towards regionalisation of military forces reflected in the appearance of stratēgiai in sigillographic sources. Starting from the reign of Leo III, according to the sigillographic documentation, the nomenclature of the provinces of late antique origin disappeared, replaced by the reference to a city or to districts, the stratēgiai, within which troops linked to that specific territory operated.^[46] Moreover, from the 730s onwards, the bullae of the kommerkiarioi are succeeded by those of the basilika kommerkia, which show a close relationship between the latter institution – the basilika kommerkia, whatever their precise activity may have been – and the supply of regional armies.^[47] This link is particularly evident in relation to the new military commands created in the eighth century, like those of Hellas, Cibyrrhaiots and Thrace.^[48] Recently, V. Prigent published a seal preserved in the British Library, which has so far gone unnoticed, of one (the name is not legible) β(ασιλικοῦ) σπ(αθαρίου) ἐκ πρ(οσώπου) τοῦ [θ(εο)]φ(υλάκτου) θ[έ]ματος τοῦ Ὀψικίου.^[49] Due to some internal elements, the author persuasively dates the seal to the 8th century; it thus constitutes a capital piece of evidence to prove that the appearance of themata in the language of the Byzantine administration would predate the 9th century. This series of indications suggests that, from the reign of Leo III or that of Constantine V, new forms of organizing and maintaining military forces were being explored that differed from those inherited from the late antique world. There was to be no ‘great’ administrative reform, but a policy promoted by the imperial government to make the relationship between military service, recruitment, and local society ever closer. The aim of this policy was to make greater use of the resources of the rural districts in which the troops operated in order to relieve the court in Constantinople, at least in part, of the burden of their provisioning and arming. Perhaps it was when the establishment of the new stratēgiai that new soldiers were recruited in the countryside, which gave rise to the neologism thema, meaning the military role on which the troops settled in a given region were registered. This would be in tune with the increasing diffusion of the term adnoumion (in the meaning of periodic review of troops by commanders) that is recorded from the eighth century onwards.

Whether the regional armies began to be constituted, as the Life of St Pancras suggests, by small, selected city troops along with soldiers settled in the territory who were mobilised on the occasion of specific operations, it is understandable that the high commands wanted to periodically check their equipment and training. Unfortunately, the state of the documentation does not allow us to do more than speculate in relation to the eighth century. However, it is significant that both the above-mentioned letter of Theodore of Stoudios and Theophane’s Chronographia seem to refer to the existence of family units responsible for hereditary military service. In the Taktikon Uspenskij, written between 812 – 813 (as Živković thinks)^[50] or between 842 – 843 (as Oikonomides thinks),^[51] we find several dignitaries mentioned, such as eparchoi, ek prosōpou, spatharioi, chartoularioi and droungarioi who are qualified τῶν θεμάτων,^[52] in addition of course to the list of stratēgoi of the themes themselves, no less than eighteen districts. Therefore, if the term thema actually originally meant ‘register’, in this administrative source of the first half of the ninth century – as in Theophanes, moreover – it has already passed to indicate, by synecdoche, the circumscription to which the registration referred and, therefore, the army itself that was stationed in it. The Taktikon Uspenskij also mentions the λογοθέτης τοῦ στρατιωτικοῦ,^[53] the official of the Constantinopolitan ministries who should have kept the antecessor of the palaios typos mentioned in the following century by Constantine VII Porphyrogenitus.

2 From the eighth to the tenth century: a perspective

In principle, military service even after the age of the Principate remained voluntary in the late Roman world. However, this principle was in fact attenuated by two trends. The first was the hereditary nature of military service for veterans’ sons, sanctioned since Constantine, but in fact probably already in existence beforehand;^[54] the second was the compulsory conscription of Roman citizens, perhaps instituted by Diocletian, and practised at least throughout the fourth century.^[55] According to this system, the richest landowners were obliged to provide one or more recruits, while small landowners were obliged to form consortia, called temones or capitula, and to bear the cost of one recruit collectively. A provision of Valens in 375 tells us that the cost of a recruit totalled 36 solidi^[56]. After the mid-fifth century, it is widely accepted among scholars that conscription in the army was mainly voluntary, except for the category of the limitanei who seem to have been obliged to serve by inheritance.^[57] Whether this situation continued during the seventh century is highly uncertain.^[58] Our evidence suggests that from the second half of the seventh century onwards the connection between military forces and local recruitment became stronger.^[59] Clear indications of the close relationship which, during the eighth century, had been established by then between private land ownership and military service are represented – certainly not by chance – by two witnesses dating back to the reign of Leo III and his son Constantine V as co-regent (March 720 – June 741). These are an article in the Ecloga regulating the division of goods of two brothers living in the same property (one of whom is a member of the army) and a sentence, issued by Leo III and Constantine V concerning a particular case in which the availability of a soldier’s assets needs to be regulated.^[60] They have already been widely discussed by scholars, making a new analysis unnecessary.^[61] Suffice it to point out that both texts show that the soldier’s maintenance in the mid-eighth century depended partly on the income provided by the family in which he lived. The judgment issued by Leo III and Constantine examines the case of husbands (γαμβροί) who, being already soldiers, become part of an oikos, bringing to it their own rhogai and physical labour. In the case of marriage failure, the stratiōtēs will have the right to keep for himself all the earnings from imperial donations, from the spoils of war and from his own salary (rhoga), while the father-in-law will have the right to claim the expenses made for the soldier’s armament (στρατιωτικὴ ἐξόπλισις), his maintenance in service (δαπάνη) and clothing (φορεσία). We do not know why the son-in-law had recourse to the financial support of his father-in-law’s property (and not to his own property) to support himself in the service. The most obvious answer is that the son-in-law did not have enough assets of his own to enable him to be a soldier. But this implies that in the mid-eighth century a patrimonial basis – whoever possessed it – was required to perform military service. And it was precisely this situation, i. e. the use of family assets to finance the military profession of a person who might not even be the owner, that gave rise to Leo and Constantine’s measure to clarify a recurring social case. Since this situation is new compared to Late Antiquity it must have been the premise for the establishment of the oikoi stratiōtikoi registers.

How did the stratiōtikoi oikoi get registered? This question implies reflection in at least two directions: (a) was enrolment voluntary or compulsory? (b) what did the enrolment procedure consist of? Once again, unfortunately, one can only make speculations on both questions. In the age of Leo VI voluntary service would have given the soldier’s household the opportunity to pay only the basic tax (and no other extraordinary forms of taxation), as well as providing the soldier himself with other income than just agricultural produce, such as imperial donations and spoils of war.^[62] Privileges deriving from military status were well known in Late Antiquity, as earnings from service in the army could allow the acquisition of (among other things) landed property and slaves.^[63] On the other hand, if military status offered privileges before tax authorities and a distinctive social condition, it also implied serious risks for the stratiōtēs, especially in the context of provincial community life between the seventh and ninth centuries. If the comparison with the late Roman world may be helpful, it should also be noted that both the hereditary nature of the service and the demand for recruits from landowners served to ensure a constant level of recruits. In a militarily turbulent period such as the fourth century, voluntary enlistment alone would probably not have been sufficient to meet the need for soldiers. Some constitutions of the Theodosian Codex testify to the phenomenon of self-mutilation of the fingers of the hand by conscripts in order to escape military service (e. g. CTh. VII 13, 4 issued in 367). It is true that in the Eastern Roman Empire, from 530s onwards, there was a militarisation of social life no different from that of the fourth century; and, even in this context, many scholars believe that enlistment remained voluntary. But the territorial framework of the sixth-century empire cannot be compared with that of the early eighth century, when the empire had shrunk dramatically and had largely lost an important recruitment reservoir that had been crucial in late antiquity, the Balkan area. At the beginning of the ninth century, both Nicephorus I (802 – 811) and Charlemagne (800 – 814 as emperor) issued strikingly similar measures, aimed at involving the freemen who did not own land in military service, by promoting consortia capable of keeping one of their members in the service.^[64] From the Carolingian capitulars it emerges that even a society, such as the Frankish one, in which the profession of arms had traditionally been associated with the notion of the freedom, denounced at the beginning of the ninth century a tendency on the part of all possessors to evade the call to arms, the heribannum. Regarding Nicephorus’ measure of 809/810, it would seem rather bizarre that the emperor expanded the numerical base of the army by forcing the economically weaker elements of the village communities to serve if the possessors with the greatest fortune had not already been called upon to bear the military burden.^[65] It is therefore not implausible to suppose that between the seventh and eighth centuries some form of compulsory conscription was imposed in Byzantium. Both article XVI, 2 of the Ecloga and the mentioned sentence of Leo III and Constantine V seem compatible with such a conclusion. I also wonder whether an obscure passage by Theophanes should not be understood in this sense. When he speaks, in an equally obscure way, of the capitation tax imposed by Leo III on Sicily and Calabria and of the seizure of the papal patrimony in those regions, he adds a piece of information not necessarily related to the situation in southern Italy. He writes that Leo III ordered (sc. the imperial officials) to watch over and register all male infants, as Pharaoh had done with the Jews.^[66] The comparison with the Egyptian Pharaoh of the Old Testament (Ex 1: 15 – 21) is not obvious – at least to me. What was the point of registering all the empire’s newborn males? Be as it may, in his Taktika emperor Leo VI uses a curious expression, speaking of τὸ λεγόμενον στρατιωτικὸν θέμα.^[67] It seems to imply that, among all the conscripts of the same thema, only a part was considered a real fighting force, of the order of 4.000 men. Moreover, Leo himself exhorts the general to choose the most valiant soldiers with economic means for his expeditions.^[68] If these indications do not clarify to what extent the registration in the stratiōtikoi oikoi was voluntary or not, they certainly suggest that between the late ninth and the early tenth century the provincial class involved in military service was strongly differentiated as far as its socioeconomic condition was concerned.

The second crucial question is how the registration of the soldiers was conceived, i. e. whether it contained the name of any male individual fiscally belonging to a certain village community (χωρίον) or, instead, the oikodespotēs, the person in charge of an oikos. In the judgment pronounced by Leo III and Constantine V the stratiōtēs is the son-in-law and the head of the household his father-in-law.^[69] It should be noted, however, that in the text of the judgment there is a passage whose translation is not without difficulty. In listing the property to which the father-in-law will be entitled after his son-in-law-soldier leaves the family, the ruling reads as follows: ‘προδήλως τοῦ πενθεροῦ καταγραφομένου ἅσπερ ἐποίησεν εἰς αὐτὸν ἐξόδους, τουτέστιν κτλ.’ (recension A); προδήλως τοῦ πενθεροῦ καταγραφομένου ἅσπερ ἐποίησατο εἰς αὐτὸν ἐξόδους, τουτέστιν κτλ (recension B). Simon, not without some doubt, interprets the passage as meaning that the father-in-law will be entitled to the expenses made for the maintenance of his son-in-law after declaring them in a public record.^[70] I wonder, however, whether the participle καταγραφομένου is not to be understood, in the active sense, as the act of registering by the father-in-law, but, in the passive sense, to indicate the condition of the father-in-law ‘who is registered’. If the passage could be understood in this way, it would be a strong indication that the stratiōtikoi kōdikes registered the heads of households, even though they might delegate the service to a family member or relative. The above-mentioned hagiographic sources of the ninth century also suggest that this second possibility – inscription of the oikodespotēs – was the most likely.^[71] Among the lexemes denoting family in ancient Greek (γένος, οἰκογένεια, συγγένεια, οἰκία, οἶκος) the latter is probably the one farthest from the strict meaning of a nucleus of blood relatives living under the same roof. In fact, according to a model dating back to Xenophon’s Oikonomikos, in the social semantics of the Byzantine world the concept of oikos not only expressed a family and its house, but also all its goods, these meanings being inseparable from each other.^[72] This implies that the owner of an oikos, after his registration in the military registers, implicitly assumed the financing of the service on his family property, whether or not it was registered. In fact, to think that to become stratiōtēs it was enough to enlist without providing any patrimonial guarantee seems unlikely when considering that the problem of the Constantinopolitan government from the second half of the seventh century onwards was to look for alternative forms of financing the army compared to the mechanisms in use in Late Antiquity, where the prefect’s administration bore the entire cost of maintaining the soldiers. The stability and transmission over time of a stratiotic oikos should have been favoured by the fact that it was both a family unit and a bundle of assets, a social cell in which the oikodespotēs could be replaced in service by sons, brothers, nephews, cousins, or sons-in-law. While the military registers were kept by the chartoularoi of each thema, the ἐπόπται τῶν θεμάτων (mentioned in the Klētorologion of Philotheos)^[73] had the task of periodically inspecting the districts for which they were responsible in order to update the changes in the assignment of the strateia.

But if the assumption of military service implied, in fact, both a personal and patrimonial burden, it seems natural that, with the passing of time, the properties of the oikoi stratiōtikoi were the object of attention of the public power. The reference to the head of the household and, at his death, to whoever within his family had replaced him as a soldier, conferred on the oikos assets a high degree of volatility in relation to their availability. It is clear that the mobility of these assets could be interpreted both in a restrictive sense (all oikos assets cannot be disposed of because they serve the military obligation) and in a concessionary sense (all oikos assets can be disposed of except those managed directly by the person who actually serves as a soldier) in the absence of an explicit mention of them in the military registers. In short, their transmission would certainly have had to deal with particular situations such as the possibility of their being acquired by a ‘political’ family – to use the language of the tenth century – through the dowry of the female branches of the military lineage, or that more than one person might find themselves exercising the obligation for families with fragmented property, or that a ‘military’ family might weaken economically and was therefore no longer able to fulfil its military duties.

In fact, this is exactly what we find in some witnesses stretching from the ninth to mid-tenth century, when Constantine issued his famous law on stratiōtika ktēmata.^[74] The transmission as dowry of a military property is explicitly mentioned in a deed of Bari from 1017, the core of which, however, dates back to the end of the ninth century, when the Byzantine Empire retook territorial control over Apulia.^[75] The division of a property is documented in another Apulian deed drawn up in Conversano in 980,^[76] which provides an example of the division of service between two brothers. A third case is contemplated in the institution of the syndotai, mentioned by Constantine VII in the De cerimoniis, namely the possibility of offering ‘contributors’ to those stratiōtai who did not have enough resources to support their service themselves.^[77] The latter practice originated in Late Antiquity and was resumed by Nicephorus I in 809/810, as mentioned above. The sources mention, therefore, not only natural ‘families’, but also increasingly larger ‘fiscal’ families contributing to the fulfilment of the burden.

It is the varied destiny of the transmission of stratiōtikoi goods, together with the tendency to their loss of economic value, that we must imagine preceded the famous law of Constantine VII of 949. This law regulates the regime of the property of the stratiōtikoi oikoi, a regime that, as the same provision explicitly states in several points, already existed previously throughout the empire, although with customs that varied from region to region.^[78] The core of the Constantinian measure is, as is well known, to determine for them a minimum economic value below which these possessions could not fall for the purpose of fulfilling the burden: 288 nomismata for knights and sailors of the themes of Aigaion Pelagos, Samos and Cibyrrhaiots; 144 nomismata for mariners serving on the imperial fleet.^[79] Historiography has generally interpreted this measure as an act breaking with the previous tradition, since military service appears in it much more markedly connoted as a patrimonial burden, rather than as a personal one. However, this ‘oppositional’ interpretation does not seem adequate for understanding the concept of oikos stratiōtikos, which from the outset must have implied a profound interrelationship between the personality of the service and its patrimonial basis, since the one did not exist without the other. The separation between these two aspects, which appears to take shape from the beginning of the eleventh century, with a progressive fiscalisation of the strateia – its fulfilment through a monetary contribution and not through an effective service – marks, in fact, an overcoming of the recruitment structure of provincial armies constituted during the eighth century. The emergence of the stratiōtika ktēmata in the legislation of the Macedonian emperors was nothing more than an acknowledgement, from an institutional point of view, of an ambiguity. It consisted in the fact that the modalities of relation between the service and its financing rested, between the eighth and ninth centuries, on a vocabulary that was inadequate from a legal point of view. The ambiguity, in fact, was rooted in the polysemy of the concept of oikos, family unit, home and patrimony at the same time. While such a notion allowed the recruitment in thematic armies to function flexibly through the wide network of the family’s cognate and agnatic structure, precisely because the oikos was also a set of economic interests, it exposed the landed patrimony of military families to splitting and erosion. Their members on the eve of Constantine VII’s law presented a very differentiated socio-economic composition. We find stratiōtai who greatly exceeded the limit of assets set by law for the performance of their service; others who were below it; still others who were completely poor. If in the eighth century the determination of property to be included among the stratiōtai could be defined by the notion of oikos, in the middle of the tenth century, at a time when the empire’s military machine was speeding up, this notion needed to be specified in its patrimonial content. Soldiers’ families were economically very uneven, and the social structure of the Byzantine countryside was populated by a multiplicity of actors – magnates, dignitaries, metropolitans, bishops, and hegumens, as well as high officials – who put pressure on the stratiotic class. This phenomenon forced the imperial power to regulate in detail the patrimonial content of the service both to strengthen socially the class of thematic soldiers^[80] and to give them greater military efficiency.^[81] In the meantime, it was from the military class of the Anatolian provinces that a number of great families acquired notoriety and prestige, such as the Argyroi, Phocas and Maleïnoi, which, in the case of the first two, would lead to their members ascending the imperial throne.^[82]

If one is legitimated to draw conclusions from such a poor framework of historical evidence, it could be said that land and military service in Byzantium during the early Middle Ages constituted a pair that developed in silent transformation. This silence is perhaps not only due to the lack of documentation, but also to the great pragmatism that guided the Isauric emperors in organising the presence of the army in the provinces and the way it was recruited. The changes in the political geography of the empire during the seventh century dictated a line of intervention aimed at consolidating the link between the control of the territory and the provisioning of the army with local resources. One of the ways of achieving this was the voluntary or compulsory enrolment of landowners and their oikoi (families and property at the same time) to promote an unattainable ideal of self-sufficiency of the soldier. This aim never succeeded in creating a virtuous economic circle, whereby local societies could finance the military class only from their own resources. The imperial power always continued to intervene, either through partial supplies of foodstuffs, donations and recruits to the thematic forces or through their support with the Constantinopolitan tagmata. But, nevertheless, this line of intervention is sufficiently clear, even in the lack of documentary evidence that afflicts the “Dark centuries”. What seems equally clear, however, is that the creation of the thematic armies and the conscription of troops in the provinces were two sides of the same coin, i. e. of the same political direction. By the 840s at the latest, if not twenty years earlier, somewhat less than twenty themata already constituted districts garrisoned by an equal number of military forces enjoying an articulated administrative organisation structure, which cannot have been created overnight, especially considering the sticky nature of the Byzantine bureaucracy. Several historians think that during the second half of the eighth century (or at the latest at the beginning of the ninth century) a new administrative district structure was established in the empire based on the recruitment areas of the army corps. If one agrees with this, it is not clear why one should think that only in the middle of the tenth century a close relationship between land and military service would have been established. In fact, very strong hints exist that even in the eighth and ninth centuries private property partly financed the maintenance of the stratiōtai. Moreover, Constantine VII’s own provision unambiguously states that such properties existed even before his measure. Historiography must rethink the separation that, in the scholarly production of the last half-century, has marked the birth of the thematic system and the patrimonial bases on which the soldiers’ service was founded. The two processes proceeded in parallel. So much so that when the themes became smaller and smaller districts, as evidenced by the Taktikon of the Escorial, new larger military districts were created, such as the Duchy of Antioch.^[83] This also prefigured a new system of recruitment during the eleventh century, with an increasing fiscalisation of the strateia and the employment of professional soldiers.^[84]