Byzantine military effectiveness against the Arabs was mixed. The imperial government found no sure means of checking or reversing their early territorial gains, and there is no evidence to suggest that any major administrative measures to redress the problem were taken specifically between 659 and 662.39 The very ease with which Mu‘awiya’s forces penetrated Anatolia in the mid-650s indicates that, in the first fifteen years following the early Islamic conquests, the government in Constantinople failed to mount effective resistance against the Muslims on the Anatolian plateau. Of events in 653/4, Sebeos writes: ‘When he [Mu‘awiya] penetrated the whole land, all the inhabitants submitted to him, those on the coast and in the mountains and in the plains.’40 Mu‘awiya’s armies were able to range far and wide, devastating Anatolia, and they could hardly have achieved this level of military activity had an effective Byzantine defence system been fully in place then. By the end of the seventh century, both states found it necessary to tighten control over the frontier zone, leaving no scope for the local populations to decide on their orientation for themselves. The Muslims even gave up the policy of allowing Cyprus to remain independent during the reign of Caliph ‘Abd al-Malik (685–705), although they soon reversed this particular decision,41 and it was difficult for any region close to Syria to maintain neutrality between the two powers. The two central governments could either introduce garrisons, as the Muslims reportedly did for a while in Cyprus, or they could evacuate the entire local population from a border zone and destroy what was left of the cities, as was the fate of Arabissos. None of these acts created hermetically sealed borders, but they did help enhance the manipulative powers of the empire and caliphate, and neither polity wanted independent buffer states to emerge between Byzantium and Umayyad Syria. Despite reports of Byzantine mobilisation during Mu‘awiya’s caliphate, it is highly unlikely that the Byzantines could have managed major military expeditions reaching into Syria. They could and did threaten Germanikeia and Melitene, and they used the Mardaites as valuable allies or surrogates, even as far afield as Lebanon. But they lacked the means and the resolve to attempt the reconquest of Antioch or other major strongholds in northern Syria, such as Chalkis, let alone any points further south. It is unclear how quickly the Byzantines’ familiarity with conditions in Syria faded after their withdrawal from there in the later 630s. A tradition has it that when CaliphMu‘awiya was informed of a string of calamities – one of his governors had run off, various prisoners had escaped and the Byzantines were raising a fresh army – the commander ‘Amr bin al-‘As advised him not to worry: ‘This is not much [trouble] for you. As for the Byzantines, satisfy them with a few concessions with which you can restrain [dissuade] them . . . AndMu‘awiya followed his advice.’42 This may be a hostile tradition, intended to malign the allegedly easy-going ways of the Umayyads, but it may also reflect a general sense among the Muslims that Byzantine threats did not need to be taken too seriously; that it was possible to reach negotiated settlements with them, without resorting to arms. The abortive rebellion of Saborios, strat¯egos of the theme of the Armeniakoi, illustrates the benefits to both empire and caliphate of direct diplomacy between Constantinople and Damascus, and Mu‘awiya’s response neatly sums up Muslim strategy in the face of Byzantine internal strife. The well-publicised failure in 667/8 of Saborios’ rebellion – for all his negotiations with Mu‘awiya43 – underlined the terrible fate awaiting those Byzantines who attempted private or personal diplomacy with Damascus. According to the chronicler Theophanes: the general of the Armeniakoi, Saborios – who was of the Persian race – rebelled against the emperor Constans. Saborios sent his general Sergios to Mu‘awiya, promising to subject Romania to Mu‘awiya if he would ally with Saborios against the emperor. When the emperor’s son, Constantine IV, learned of this, he sent Andrew the koubikoularios to Mu‘awiya with gifts so that he would not cooperate with the rebel. Mu‘awiya reportedly declared: ‘You are both enemies, I will help him who gives the most,’ to which Andrew replied: ‘You should not doubt, caliph, that it is better for you to get a little from the emperor than a greater deal from a rebel.’ Although the revolt enabled the Muslims to capture Amorion, the administrative centre of the Anatolikoi theme, and to raid as far as the Bosporus, the Byzantines soon seized the city back, annihilating the Muslim garrison that had been installed there.44 Saborios’ revolt marked a high point in Umayyad diplomatic attempts to win control of the Byzantine empire through negotiations with the local Byzantine commanders. The Muslims hoped to peel away segments of the empire by convincing local Byzantine (or Byzantino-Armenian) border commanders to break away, perhaps to found neutral buffer states or even to switch allegiance outright, allowing theMuslims to occupy these border areas and raise tribute from them. The miserable fate of Saborios and his supporters reinforced imperial authority, strengthening the belief that revolt against Constantinople or direct negotiations with theMuslims would only result in death and destruction.45 Despite a few early, encouraging examples of local Byzantine towns in Syria and Egypt surrendering to theMuslims, this did not become a trend.46 WhileMu‘awiya hoped to exploit tensions between Greeks and Armenians on the Byzantine side of the frontier, Constantinople employed a range of policies and techniques to enforce the emperor’s authority there. These included appointing skilful and ruthless eunuchs to punish and put to death anyone who attempted to become separatists, or who toyed with coming to terms with the Muslims on their own. Constantinople’s efforts paid off: the core areas of Byzantine Anatolia lacked commanders who would find it in their best interests to switch sides between Constantinople andDamascus.Mu‘awiya and his successors failed to find a single strat¯egos or senior officer within the all-important theme of the Anatolikoi who would be willing to betray his command to theMuslims. The best-known example ofMuslim attempts to subvert a Byzantine border commander are the negotiations in 717 between the commander-in-chief of the great expedition against Constantinople, Maslama bin ‘Abd al-Malik, his field commander, Suleiman bin Mu‘ad, and Leo ‘the Isaurian’, the wily strat¯egos of the Anatolikoi. Leo reportedly parleyed with Suleiman for several days near Amorion. But for all his show of readiness to offer tribute and even reportedly to discuss with Muslim emissaries ways of handing the empire over once he had ensconced himself in Constantinople, Leo never intended to submit to the Umayyads: his was a long-drawn-out ruse, as Suleiman and Maslama learned to their bitter regret. These negotiations helped Leo to gain the throne, but they brought only embarrassment and defeat to the Muslims.47 The utmost care was taken by the emperor in selecting commanders of the theme of the Anatolikoi. This was the most powerful field command, and despite occasional rebellions, the strat¯egoi of the Anatolikoi never betrayed their commands to the Muslims. Had they done so, the overland road to Constantinople would have lain open to the enemy. Although Umayyad Damascus and its court continued to hope for such an opportunity, it eventually became apparent that the problem of Syria’s northern borders would not be resolved by Byzantine commanders’ switching sides. The empire proved resilient, as it restored a degree of control over its borders and peripheral regions. There was also an inherent contradiction between the desire of some Muslims to amass booty for themselves from Anatolian raiding and Damascus’ need to reach a modus vivendi with the local inhabitants and leaders in the border regions. Greek and Roman military maxims shaped how the Byzantines saw the warfare against Muslim Syria, and it is unclear how successfully they digested their own, much more recent experience of military catastrophe there. Until about 711 the reigning Heraclian dynasty may well have made it awkward for anyone to offer a written historical analysis of events. Such inhibitions would have eased from 711, but by then Byzantine Syria had more or less passed from living memory, except among a small number of renegades and refugees. Despite recent warfare, the borders were now gaining durable, albeit still uncertain, parameters. As with the caliphate, the empire suffered from acute internal rivalries, discouraging the emperors and their advisers fromgiving adequate resources or total confidence to the best military commanders. There was a deepseated fear in Constantinople that well-resourced generals might be able to exploit newly won military victories to overthrow the government. As long as the empire’s Armenians were less than reliable in their loyalties, any long-term offensive against Umayyad Syria was impractical, no matter how much money Byzantium might extort from Damascus under fitnainduced truces. Only the Armenians could provide enough hardy military manpower for the Byzantines, yet the imperial government’s relationship with the Armenians living in Caucasian regions underMuslim control was ambivalent and many-stranded (see above, ch. 8). Finally, it is worth noting that seventh- and earlier eighth-century Byzantines andMuslims lived in a mental environment of eschatological, indeed apocalyptic expectations, although they were not explicitly linked with the approach of any specific millennium. Those fears and hopes remained strong throughout the seventh century and were to be found in many regions, both east and west. They affected and nurtured a number of religious manifestations and movements within Greek, Armenian, Syriac monophysite and Muslim communities. Apocalyptic expectations soared in the middle of the seventh century, perhaps peaking in the reign of Leo III (717–41), as the centenary of the appearance of Islam approached.