1The artisanal production of second Iconoclasm has not generated much discussion since Andre Grabar’s L’iconoclasme byzantin was first published in 1957,1 though considerable evidence survives, in the form of actual objects and descriptions of them. The context within which that production must be understood is also reasonably clear. From a material point of view, it consists of three bodies of objects.

THE CONTEXT FOR THE ARTISANAL PRODUCTION OF SECOND ICONOCLASM

2The texts preserved from between ca 600 and ca 900 in Byzantium attest to the force of the past in early medieval Byzantium: tradition had been a major point of reference since at least the advent of Islam, and remained important throughout Iconoclasm. The aspects of the past that iconoclasts presented as part of their rhetoric of self-validation were not the same as those used for the same purpose by the iconophiles, but iconoclasts and iconophiles alike called upon tradition to justify systems of belief that each side believed to be quite different from that held by their opposition.2 Since different aspects of the past were valorised by each side, this did not mean that the old was automatically venerable - even outside of polemical debates like Iconoclasm, it is clear that this was not necessarily the case in Byzantium – but it did mean that a particular aspect of the past that was promoted by a special interest group was hard for friend and foe alike to ignore.

3Religious images, debated as a proposition during Iconoclasm, fit this profile. Precisely because religious portraiture was under scrutiny, the context – the ambience - created by works made before first Iconoclasm that had survived until the beginning of second Iconoclasm in 815 could not be ignored even by those who were active iconoclasts;3 for those who passively continued to accept the significance of religious art, and especially for those who actively defended it, the presence of works sanctified by time was presumably even more significant. Some of these works (for example, the encaustic icons now housed at the Monastery of StCatherine on Mount Sinai) were held in areas that were outside the area directly affected by Iconoclasm.4 Others – such as the sixth- and seventh-century mosaics at Hagios Demetrios in Thessalonike -were well within the territories under imperial Iconoclast control, but they nonetheless outlived the controversy. There is no evidence that the mosaics and frescoes at Hagios Demetrios were covered up,5 and the little documentation we have suggests that even in Constantinople (the bastion of imperial Iconoclasm) some mural decoration survived Iconoclasm intact.

4It is worth stressing that the Byzantines themselves rarely used the word Iconoclasm; instead, they preferred iconomachy (image struggle), a term that more accurately responds to the period as we now understand it: there is remarkably little evidence for any actual iconoclast destruction;6 and, as the often-cited letter sent by the emperors Michael II and Theophilos in 824 to Louis the Pious admits, the iconoclast emperors continued to allow ’those images that had been placed higher up to remain’.7 This is important not only because it demonstrates that Byzantine Iconoclasm was more flexible than later, European, iconoclast movements, but also because it means that the context for works produced during Iconoclasm was predetermined: anything produced during Iconoclasm had to be seen within a framework created long before it, and in a framework that was not, on the whole, visibly affected by Iconoclasm.

5A second context for works produced during second Iconoclasm was, obviously, that generated by works produced during first Iconoclasm (i.e. in the half century between ca 730 and 787). Since second Iconoclasm was in some ways a conscious restitution of those earlier policies, one would suspect that products of first Iconoclasm might be particularly valued or emulated. It is, however, difficult to evaluate this possibility. For example: Leo III and Constantine V were both responsible for considerable repairs to the urban fabric of Constantinople during first Iconoclasm, a process continued by Theophilos during second Iconoclasm, as exemplified by his extensive rebuilding of the city walls.8 But Theophilos did not claim that his wall restorations continued the policies of his glorious forebears, though he inserted many inscriptions on the walls where he might have done so: his project must be counted as continuity driven by necessity rather than by a desire to emulate.9

6A more obvious candidate for the emulation of first Iconoclasm by patrons and artisans of second Iconoclasm is the decoration of church apses with large crosses. This motif is first preserved in the capital in Hagia Eirene, damaged in the earthquake of 740 and rebuilt and redecorated in the midst of first Iconoclasm, shortly after 753.10 Unfortunately, however, other examples are less clearly dated: we do not know when the cross decoration was placed in the apse at the Koimesis church in Nicaea,11 while the apsidal cross in Hagia Sophia at Thessalonike appears to belong to the considerably later programme of decoration commissioned by Eirene and Constantine VI sometime between 780 and 797 - that is, it is not necessarily iconoclast at all.12 The only building that may probably be securely ascribed to second Iconoclasm that assuredly follows this tradition is Hagia Sophia at Vize in Thrace, dated dendrochronologically to after 833, which incorporates crosses sculpted in relief and has painted crosses in the apse.13 While no other monuments with aniconic decoration prominently featuring the cross are independently dated,14 it nonetheless seems intrinsically likely that apsidal cross decoration was continued during second Iconoclasm under the direct inspiration of first Iconoclasm.

7The only other instance where there is an incontestable impact of first Iconoclasm on the products of second Iconoclasm appears in the coinage of Theophilos. Here, the practice of depicting a portrait of the deceased former emperor on the reverse - instituted in Byzantium (probably following Islamic precedents) by Constantine V, who placed his father, Leo III, on the reverse of some of his coins - was revived on the ’class III’ nomismata minted under Theophilos between 830 and 840, which portrayed the dead Michael II on the reverse.15

8The third context for the products of second Iconoclasm can be treated very briefly. This is the body of works produced between the end of first Iconoclasm in 787 and the beginning of second Iconoclasm in 815. Excluding religious figural imagery such as encolpia engraved with Christian scenes and textiles like the Vatican Annunciation and Nativity silk,16 which are unlikely to have been emulated in iconoclast Constantinople, we are left with the aniconic mosaics commissioned by Eirene and Constantine VI for Hagia Sophia at Thessalonike (mentioned earlier) and the votive panels donated by Eirene to the Church of the Virgin of the Source in Constantinople themselves continued earlier patterns.17 Both of these in fact continued earlier patterns, and cannot be seen as inspiring subsequent work. Eirene's one apparent innovation' - the portrait of Christ on the Chalke Gate - pretended to be a restoration; and, according to the Scriptor Incertus, Leo V duly removed it at the start of second Iconoclasm in what he thought was direct emulation of Leo III.18 Again, it is in the realm of numismatics that a pattern initiated earlier recurs in second Iconoclasm: the double portrait found on the nomismata of Eirene, where she appears on both the obverse and the reverse, was copied by Leo V and Michael II (and perhaps Michael I).19

ARTISANAL PRODUCTION OF SECOND ICONOCLASM

Portable goods

9Two manuscripts well-known to Byzantinists date from the period of second Iconoclasm,20 but neither contains miniatures or elaborate ornament. Paris, gr. 437 is a majuscule copy of pseudo-Dionysios that was probably written shortly before 827 and sent west by Michael II as a gift to Louis the Pious.21 The manuscript incorporates simple decoration: small red crosses accompany chapter headings, and some enlarged initials are elongated. To this basic scribal repertoire, the Uspenskij Psalter (Petropol. gr. 219) - a minuscule text written by the Stoudite scribe Nicholas in 83522 - adds a small terminal ivy leaf, drawn in the ink of the text, to the base of some marginal scholia. The St Petersburg manuscript is well-known as an early example of minuscule, and as an early product of the Stoudite monasteries; it should also be recognised as one of the oldest Byzantine manuscripts to incorporate what would soon become the ubiquitous ornament of ninth- and tenth-century Greek manuscripts: its ivy leaf terminal decoration will dominate the decorative systems of all Greek books produced for the next century.23

10Coins and seals aside,24 the most significant metalwork produced between 815 and 843 was the 'Beautiful Door' at Hagia Sophia in Istanbul.25 The door is composed of copper-alloy plates attached to a wooden core; the two panels measure 4.35 metres in height and, together, 2.91 metres in width. It was installed in 838/9 and its legends were revised in 840/1 to accommodate the birth of Michael III. Presumably in deference to the medium, the inscriptions mimic those on contemporary seals; they read in translation 'Lord help the ruler Theophilos', 'Mother of God help the empress Theodora', 'Christ help the ruler Michael', 'Theophilos and Michael, victors', and give the date. The framing strips are decorated with acanthus scrolls, foliage, a meander pattern, and a scrolling pattern filled with grape clusters, fleur-de-lys, and five-lobed leaves. These are familiar ninth-century decorative motifs, all of which recur in, for example, manuscript decoration of the second half of the century.26

11In the mid-tenth-century Book of Ceremonies, the 'Beautiful Door' is one of the main imperial entrances to the church, and, at about the same time as this text was written, the lunette above the door was decorated with a mosaic showing Constantine the Great and Justinian presenting Constantinople and Hagia Sophia, respectively, to the Virgin.27 Theophilos's door suggests that the importance of the entrance dates back to the years of second Iconoclasm, during which period the southwest corner of Hagia Sophia apparently under-went reconstruction.28

12Other metalwork associated with Theophilos has not survived. This included 'two enormous organs of pure gold...decorated with different stones and glasses', a 'golden tree in which were perched birds that warbled musically by means of some device', and, from the mint, the Pentapyrgion, a large display case housed in the Chrysotriklinos (throne room) of the imperial palace.29

13Icon and textile production is more problematic because there are no examples firmly dated to the years of second Iconoclasm. Icons probably continued to be made, at least in areas away from the capital,30 but evidence for Constantinople is lacking. Authors claiming that female members of the imperial household venerated holy portraits in secret wrote a century or more after the Triumph of Orthodoxy, and the only contemporary source praising an empress for rejecting Iconoclasm - Theodore of Stoudion's letter to Theodosia, widow of Leo V († 820) - does not mention icons.31

14Textiles, too, apparently continued to be produced. Michael II sent ten silks to the Carolingian emperor Louis the Pious, and Leo Grammatikos tells us that the imperial vestments were renovated and 'adorned with gold embroidery' during the reign of Theophilos.32 Conclusive identification of actual fabrics produced during the years between 815 and 843 is not, however, possible. Silks such as the Aachen and Brussels charioteers, in the past sometimes associated with second Iconoclasm, have recently been more broadly dated to the eighth or ninth century.33 Other pieces can be tied relatively firmly to western owners contemporary with second Iconoclasm - the Vatican Pegasus silk, for example, is associated with the cross of Pope Pascal I (817-824)34 - but because silks were kept in store or reused it is impossible to know for certain that it was made during Pascal's papacy or beforehand. Subject matter is still sometimes used as a dating criterion: the London Charioteer silk has, for example, been associated with Theophilos, who, probably as part of his victory celebrations after the pillage of Zapetra in 837, participated in (and of course won) a race in the Hippodrome;35 and it has been suggested that for silks with secular motifs 'it would be reasonable to think that [they were] woven sometime during one of the two periods of Iconoclasm... when figurative religious subject matter was effectively banned'.36 But since 'secular' motifs survive in silks of many periods, this argument is not persuasive, and there are, in the end, no silks that can conclusively be tied to the years between 815 and 843.

Architecture and architectural decoration

15Few monuments from the iconoclast years have been fully preserved, but what has managed to survive is sufficient to indicate the development of a new type of church plan, usually called the cross-in-square, by ca 800.37 This plan dominated orthodox religious architecture throughout the remainder of the Byzantine period, and remains the standard plan used for orthodox churches to this day. But although second Iconoclasm cannot take credit for the cross-in-square plan, it is nonetheless evident that considerable building was undertaken during the period.

16In Constantinople, the baptistery of Hagia Sophia was remodelled or repaired sometime after 814, a date provided by dendrochronological testing of one of its beams.38 This work may well have coincided with a larger remodelling of the southwest corner of the Great Church that saw the installation of Theophilos's 'Beautiful Door' at the new imperial entrance to the building, discussed above.

17In the immediate environs of the capital (including the Marmara islands), Hagia Sophia at Vize in Thrace is dated dendrochronologically to after 833; it is architecturally related to Hagia Eirene in Constantinople and, as we have already seen, includes sculpted crosses and painted crosses in the apse.39 Other structures are less well studied. The Chalki monastery on Heybeliada (the island of Theodore of Stoudion's exile) was founded by a certain John in the early ninth century; Vincenzo Ruggieri believes that he has found its ruins.40 On St Glyceria island (Incirli adasi), the monastery to which Niketas, hegou-menos of Medikion, was exiled after 815 is partially preserved but has not yet been excavated.41 Finally, the monastery of St Constantine on the island of Thasios (Monastir adasi) in lake Apolyont (Bithynia), is probably of the ninth century, and perhaps to be associated with St Ioannikios.42 These last three may or may not belong to second Iconoclasm: we await the results of excavation.

18Elsewhere,43 at Side, on the south coast of Anatolia, church H reveals an early version of the cross-in-square plan, and may probably be dated to the ninth century.44 Somewhat better preserved, a church on Söçüt adasi (Bozburun area) has been dated to the first half of the ninth century.45 A few additional sites are known through inscriptions. Just prior to second Iconoclasm, an inscription of 2 May 812 records the enkainia of a church of the archangel Gabriel at Alakilise in central Lycia,46 and an inscription of 813 records Sisinnios's restoration of the monastery of the Theotokos at Corlu in Thrace.47 During second Iconoclasm itself, an inscription now held in the Tigani museum attributes some unspecified (reconstruction?) work on the island of Samos to the reign of Theophilos.48

***

19Various texts - mostly saintly vitae - suggest that many more monuments than those just listed were built during second Iconoclasm. The documentary evidence confirms a number of patterns suggested by the preserved monuments; it is also important because it reveals themes that would otherwise be totally lost to us. Of course, these documents have narrative agendas of their own; they are not the primary material, the monument itself, but a mediated interpretation of that monument that may even, at times, describe a building that never existed. Documents on their own are thus never entirely trustworthy, and can only be used for certain (restricted) types of interpretation.

20An exemplification of the types of problems raised by descriptions of buildings concerns what is perhaps the best-known of all monuments attributed to second Iconoclasm: Theophilos's Bryas Palace, described by Theophanes Continuatus.49 Though it has been argued that the ruins of this palace survived at Küçükyali (in the Asiatic suburbs of Istanbul, across from the Prinkipo islands), Alessandra Ricci's recent work suggests that this is not the case: the ruins appear to be the remains of a cistern above which was constructed a single-apsed and domed church that apparently closely resembled the eleventh-century monastic church of St George at the Mangana.50 We must return to the continuator of Theophanes for our information on the Bryas Palace. He says that John the synkellos (later to become the famous iconoclast patriarch John the Grammarian) returned from an embassy to Baghdad deeply impressed, and

Having come back to Theophilos and described to him the things [he had seen]... [John] persuaded [Theophanes] to build the palace of Bryas in imitation of Arab [palaces] and in no way differing from the latter either in form or decoration. The work was carried out according to John's instructions by a man named Patrikes who happened to be also adorned with the rank of patrikios. The only departure he made [from the Arab model] was that he built next to the bedchamber a church of our most holy lady, the Mother of God, and in the courtyard of the same palace a triconch church51 of great beauty and exceptional size, the middle part of which was dedicated to the archangel [Michael], while the lateral parts were dedicated to women martyrs.52

21This account has provoked considerable discussion, and has been interpreted variously as evidence of cultural receptivity (or appropriation) under Theophilos,53 or as evidence of later iconophile fabrication intended to paint the last iconoclast emperor and his patriarch as Islamic sympathisers.54

22The account of Theophilos's additions to the Great Palace in Theophanes Continuatus is less controversial and less open to polemic interpretation.55 The palace constructions consisted of a number of buildings sheathed in various kinds of marble, with bronze and silver doors (which possibly resembled the 'Beautiful door' at Hagia Sophia), and a bronze fountain with a 'rim crowned with silver and a gilded cone' that spouted wine during the opening reception for the building, 'two bronze lions with gaping mouths [that] spouted water', and 'carved verses' by Stephen Kapetolites and Tgnatios, the university professor'. An armoury was decorated with pictures of 'shields and all kinds of weapons'; another room had its lower walls reveted with slabs of marble 'while the upper part [had] gold mosaic representing figures picking fruit'; a third had 'on its walls... mosaics whose background is entirely gold, while the rest consists of trees and green ornamental forms'. Theophilos is also credited with decorating 'with gold mosaic' two pre-existing halls, the Lausiakos and the hall of Justinian II; and the whole complex was interspersed with terraces and gardens.

23The Continuator spends more time describing the marble than the mosaic, but certain elements of the decoration nonetheless recall the mosaics in the Dome of the Rock at Jerusalem and in the courtyard of the Great Mosque at Damascus, by the time of Theophilos over a century old. These were probably set at least in part by imported Byzantine mosaicists - the technical details are remarkably consistent with slightly earlier Constantinopolitan work - and. along with the contemporary Umayyad palaces, suggest a range of deluxe non-religious imagery of which little has been preserved in Byzantium. There is no reason to link such decoration specifically with second Iconoclasm (earlier and later accounts of comparable ornament survive in abundance), but its apparent continuation during the second quarter of the ninth century indicates that training in 'the art of mosaic' did not wither then: artisans skilled in setting tesserae could easily have been employed through the iconoclast era.

24In and around Constantinople, imperially sponsored monasteries noted in the documents are limited to the Monastery of Gastria, favoured by the women of Theophilos's family (his mother Euphrosyne retired there, and his widow Theodora was exiled and buried there);56 the monastery of St Panteleimon erected by Theodora;57 and the monastery at ta Anthemiou, Chrysopolis (on the Asiatic shore of the Bosphoros), built by Alexios Mousele, son-in-law of Theophilos, who retired there after 843.58

25Theodore of Stoudion noted a convent dedicated to the Theotokos built by Anna, wife of Leo the patrikios; 59later sources suggest that the poet Kassia may also have founded a monastery in 823.60 The patrikios Antony apparently founded the church of the Theotokos of Neorion, probably in the 820s;61 while after his death from wounds in 838 the Armenian general Manuel was buried in a monastery that apparently formed part of his palace.62 Soon thereafter, in 840 or 842, Leo Grammatikos seems to suggest that the military leader Theophobos' house was turned into a monastery after his execution for treason by Theophilos.63 The same emperor also apparently ordered his courtier Martiniakos, in disgrace, to convert his estate into a monastery and retire there.64

26We also hear of monasteries founded outside of the city by exiled icono-philes. According to his vita, the former patriarch Nikephoros founded the monastery of ta Agathou near Chalcedon, at the first stop of his exile in 815.65 In the same year, St Makarios, also in exile, is said to have founded an eponymous monastery on the Bosphoros, after visiting Nikephoros, 'on the straights of the Propontis'.66

27The island of Aphousia (Avsa adasi) in the sea of Marmara was a common site of banishment in the ninth century. A church was built here by Makarios of Pelekete after 829.67 Also during the reign of Theophilos, Symeon of Lesbos built a church dedicated to the Theotokos here; this later became the core of a monastery.68 Further south, near the Rhyndakos (Hellespont), a monastery dedicated to St Porphyrios belonged to Peter of Atroa before 830.69

28Four monastic complexes are also signalled in various sources. For our period, the monastic complex on Mt Olympus in Bithynia is known primarily from the vitae of St Ioannikos (752/4-846) and Peter of Atroa (f 837). The best known monastery here was that dedicated to St Zacharias, apparently restored ca 800 from an abandoned chapel by Peter and his master Paul of Atroa.70 During second Iconoclasm, we are told that a church dedicated to John the Baptist was built in the nearby monastery of Antidion by Ioannikios in (probably) the 830s.71

29A second group is associated with Niketas the Patrician. The Monastery of Niketas at Zouloupas (near Nikodemia) was built between 833 and 836 on land given to Niketas by a relative; apparently, a monastic community was established here.72 After leaving Zouloupas, Niketas restored the church of St Michael at Katesia between 833 and 836.73 The Monastery of Niketas near Kerpe on the Black Sea was built before 836; Niketas was buried here.74

30Thirdly, before he became patriarch of Constantinople in 847, Ignatios (the castrated son of the former emperor Michael I) built three monasteries on islands in the sea of Marmara: at Terebinthos,75 Yatros,76 and Plate. The latter, from which a seal has been preserved, included a Church of the Forty Martyrs and a chapel dedicated to the Theotokos.77

31Further afield, a group of three monasteries on Mount Lissos (in Lydia) associated with the iconophile Ioannikios were all apparently built ca 815. The vitae of Ioannikios name them as the monastery of the Apostles (with a typikon composed by Ioannikios);78 the monastery of the Theotokos, which was perhaps joined with the Apostles monastery (here too Ioannikios left a typikon for the appointed hegoumenos); 79and the monastery of St Euphemia, near to the other two.80

32Also in Lydia, a monastery of Balentia is noted in the life of Peter of Atroa,81 and a church dedicated to St Isidore is recorded as existing on an island near Lesbos when Symeon of Mityline arrived there around 820.82 Whether either of these belong to second Iconoclasm is, however, uncertain.

33Finally, the documents note a limited amount of restoration work in the region of Constantinople. For example, a nunnery dedicated to Metanoia (independently attested by a seventh- or eighth-century seal) that had been revived already by Theodote, the second wife of Constantine VI, had become structurally unsound by 840 when, at the request of its resident nuns, it was converted into a xenon by Theophilos.83

Painting

34No representational painting can be securely dated to second Iconoclasm,84 though decorative geometrical mosaic work has recently been recovered from the Fatih Camii in Trilye (post-799) and some was also found at the Vize church mentioned earlier, which dates to sometime after 833.85 At Vize, too, fragments of wall painting were found, and first fully published in 1989.86 The largest fragment shows remnants of a composition with the enthroned Christ holding an open book on which the standard T am the light of the world' text (John 8:12) is inscribed; an angel and an unidentifiable figure, perhaps female, stand to the right. Yildiz Ötüken, who published the fragment, dated it to the later ninth century, on the assumption that representational religious imagery could not have been painted earlier in the century. I wonder. The iconography could belong in the seventh, eighth or ninth century, though it recalls Hagios Demetrios in Thessalonike more than post-iconoclast monuments largely known from Constantinople. In terms of style, the fragment is too damaged to bear the weight of much speculation, but the drapery patterns that can be minimally traced fit within a broad strand of geometricizing drapery painting that would not surprise anytime from the mid-eighth century - when it appears in the Vatican Ptolemy - through the first three-quarters of the ninth century, when it is found, for example, in the mosaics attributed to the 870s at Hagia Sophia in Constantinople.87 On the basis of style and iconography, it is possible - though no more than that - that Vize was painted when it was built, sometime shortly after 833.

35More solid evidence for representational imagery during second Iconoclasm might be extrapolated from images produced after the so-called 'triumph of orthodoxy' in 843. The miniatures of the Khludov Psalter did not emerge from a vacuum, and if Kathleen Corrigan is correct in her association of the manuscript with Methodios and its dating to 843-847, the preparation for the pointed political cartoons that are such a famous feature of that manuscript must have occurred during the reign of Theophilos.88

36Documentary evidence for figural religious imagery during second Iconoclasm appears in (later) references to the painter Lazaros, a man so shrouded in iconophile rhetoric that his legendary artisanal production cannot easily be translated into actual artefacts. What little information has been preserved about him was collected and synthesized long ago by Cyril Mango.89 Lazaros is recorded in the Synaxarion of Constantinople, the Liber Pontificalis, and a papal letter; the major account of his activities, however, appears in Theophanes Continuatus.90 This repeats many of the well-worn tropes favoured by iconophile authors, claims that Lazaros was 'famous for the art of painting', and was 'widely believed to have died' from the torture Theophilos had inflicted upon him for this fame; he had "barely recovered in prison, [before he] took up his art again and represented images of saints on panels'. Theophilos therefore 'gave orders that sheets of red-hot iron should be applied to the palms of his hands'. This almost killed him, but then, 'thanks to the supplication of the empress [Theodora] and some of his closer associates', Theophilos released Lazaros from prison and he went to the church of the Forerunner tou Phoberou (on the Asiatic shore of Bosphoros); here, according to the Continuator, he painted an image of John the Baptist that later performed 'many cures' (the exact nature of which is unspecified). Theophanes Continuatus also credits Lazaros with the Chalke portrait of Christ set up in or after 843.

37Whatever his artisanal production during second Iconoclasm actually was, Lazaros evidently did not die for it; and that he became a symbolic locus for later accounts of iconophile activity is clear from pilgrim accounts of Hagia Sophia from ca 1200, where Lazaros is credited with the apse mosaic dedicated by the patriarch Photios in the presence of Michael III and Basil I in 867.91 The person Lazaros apparently existed, but the products attributed to him cannot be tied to any preserved works.

CONCLUSIONS

38Artisanal production did not stop during second Iconoclasm. In terms of preserved works, however, no more remains from second than from first Iconoclasm and the interim period between 787 and 815. Any attempt to define the artisanal production of thirty years is bound to run into difficulties, and this summary of known works from the years between 815 and 843 is no exception. A few general points are nonetheless worth remarking.

39Much of the construction work documented during Iconoclasm consists of repairs to or alterations of older structures. There may well be more of this than the sources suggest, for it seems at least plausible that some of the monasteries that our documents claim were 'built' - like the converted estates of Theophobos and Martiniakos mentioned earlier - in reality adapted older domestic buildings for monastic use. But both repair-work and new construction alike require a skilled labour force, and this seems to have been available throughout the eighth and ninth centuries. That it was mobile is suggested by the ninth-century (but post-iconoclast) church at Dereagzi in central Lycia, the bricks for which came from a region on the sea of Marmara, probably with a master mason to lay them.92

40Land-use and patronage-patterns during Iconoclasm as a whole are also worth noting. Some land seems to have been developed for the first time, and this suggests economic expansion. The islands in the sea of Marmara apparently underwent extensive development, perhaps because during the years of Iconoclasm they served as places of exile, and were sufficiently removed from the capital to allow considerable freedom of movement (and patronage) for banished aristocrats. Many new monasteries - which, as noted earlier, may well sometimes have been little more than refurbished domestic buildings -were constructed on family lands the previous use of which (if any) is rarely specified, though occasionally, as at the Stoudite Monastery at Boskytion (in Bithynia), we are told that the complex was built on family estates already under cultivation.93

41One of the more interesting general conclusions to be drawn from the evidence preserved from the iconoclast centuries is that while iconoclast emperors are credited with civic patronage such as repair of walls and aqueducts, ecclesiastical patronage in the period is less often attributed to them than one might have anticipated. The identity of many patrons is not known, and the majority of named benefactors were religious. Others, however, were civic officials of some sort (e.g. Anna, wife of Leo the patrikios, the general Manuel, the courtier Martiniakos);94 and some were aristocrats who sponsored monasteries or groups of monasteries (e.g. Antony and Niketas, both patrikioi). This may be due to the nature of our source material, which is largely hagiographic and pro-icono-phile: the longest-ruling emperors of the years between 730 and 843 were Leo III and Constantine V (who between them ruled from 717 to 775) and, since both were vilified as heretics by later writers, positive references to any ecclesiastical building activities may have been expunged from the record. The semi-legendary account of the building history of Constantinople preserved as the Patria follows this same pattern, with few churches attributed to iconoclast emperors; here, as in many tenth-century and later revisionist accounts of the period, ecclesiastical initiative during Iconoclasm is linked to imperial wives.95 That the ninth-century written material anticipates this model suggests that the history of the iconoclast era began to be revised earlier than has sometimes been proposed.

42In terms of artisanal products, second Iconoclasm did not witness a sharp break from the past. As we saw at the beginning of this discussion, the dominant force was continuity rather than rupture. On the whole, pre-iconoclast monuments survived, and apart from the general shift away from figurative religious subject matter there is little to differentiate the patterns of production from those that came before. The most radical architectural change of the whole period was the introduction of the cross-in-square plan, but that, as we have seen, pre-dates second Iconoclasm and its development simply continued in the first half of the ninth century. The same is true of the two major technical innovations of the iconoclast era, minuscule and a new loom type: both apparently emerged ca 800 and stayed on.96 That the key date here is ca 800 suggests that the real shifts were economic ones, and had less to do with Iconoclasm (first, second or the interim between) than with the increasing prosperity of the ninth-century world. But one thing was radically changed by Iconoclasm: the roles and uses of figurative religious art. Codified and developed during the debates of the eighth and early ninth centuries, the theory of images charged icons with qualities previously held only by relics and thus fundamentally transformed the Orthodox experience.97