Travelling Fresco Painters in the Aegean Late Bronze Age: The Diffusion Patterns of a Prestigious Art

It forms a preliminary introduction to a much longer and more detailed study which will appear elsewhere in due course, covering all aspects of the emergence and spread of Late Bronze Age figured fresco painting, from its Knossian origins to its final manifestations on the Greek mainland.

INTRODUCTION

Murals, as an art form applied to and wholly dependent on architecture, by definition do not travel. As a result, a necessary condition for their transmission - that is, their first appearance in a particular locality and their subsequent inter-regional diffusion - is the mobility of painters. This is particularly true in that the practice of mural painting requires a high level of specialised technical knowledge together with artistic skill, especially in the case of figurative compositions.

With this simple, if not obvious, principle as my starting point, I shall try to approach the apparently straightforward subject of the emergence, transmission and mechanisms of selective diffusion of Aegean pictorial fresco painting in the second millennium BC. As the major form of 'public' applied art, frescoes illustrate the artistic and social codes of the ruling classes, and more generally of the elite of the era, by providing the means for a symbolic, programmatic marking of built space for religious, social and self-promotional purposes, not infrequently of a propagandistic nature (Hägg 1985; Marinatos 1985, esp. 222-229; Boulotis 1987, esp. 153-155; 1992, 90-93; 1995, 13, 21-32; Niemeier 1992; Blakolmer 1995). It is thus precisely within this socio-symbolic sphere that we must look for the rationale behind the mobility of the painters and the diffusion of their art.

The chronological frame for my observations corresponds mainly to the Late Bronze Age, while the geographical frame, with Crete as its pivot, is defined by the southern Aegean area (including southeastern Asia Minor), by Mycenaean Greece and by the eastern Mediterranean. Within this chronological and geographical frame, which is characterised by intense social mobility and transformations in material and symbolic expressions as well as by extensive cultural interaction, the history of fresco painting follows a dynamic path of quantitative increase and geographical diffusion. This path is marked by the following basic stages:  

     a)    With the construction of the Second Palaces (MM IIIA, ca. 1700 BC), pictorialisation made its appearance in fresco painting, following preliminary attempts at Knossos and Phaistos during the period of the Old Palaces (Immerwahr 1990a, 21-22, fig. 6). Although this art form effectively died out at Phaistos after the beginning of the neopalatial period, and was apparently used on a comparatively restricted scale at the other minor palaces, such as Mallia and Zakros (Boulotis 1992, 87-88; 1995, 14-15; Blakolmer 1997), at Knossos itself it experienced a quantitative as well as qualitative and thematic explosion, which led M. Cameron to talk about a Knossian "fresco industry" (Cameron 1975, 271; 1976, 12). The same occurred in the case of the related, but more prestigious and time-consuming category of painted stucco reliefs (Kaiser 1976, esp. 307-308).  

     b)    At its peak, that is during LM IA, fresco painting, with Knossos justifiably regarded as the vital centre, was widely diffused not only in Crete, but also outside the island at centres which had clearly been subjected to Minoan influence, especially in the Cyclades (Akrotiri on Thera, Phylakopi on Melos, Ayia Irini on Kea), the Dodecanese (Ialysos/Trianda on Rhodes, the Serraglio on Kos), southeastern Asia Minor (Miletus), probably already on the Mycenaean mainland and, with an astonishingly wide-reaching spread, as far as Syro-Palestine and Egypt.  

     c)    During the LM II and LM IIIA1 periods, the brilliant art of mural painting on Crete itself was confined to the palace of Knossos and to Ayia Triada, where it reached its ultimate floruit. Meanwhile, it had almost completely vanished from the rest of Crete and the Aegean islands. However, in the newly founded palaces of Mycenaean Greece (ca. 1400 BC) there began a vivacious new phase in the art of mural painting, based apparently on the Knossian model, which was to last for nearly two centuries, down to the end of LH IIIB. Humble remnants of this fresco art are still in evidence at the beginning of the LH IIIC period, in some fragments from the Cult Centre at Mycenae (Kritseli-Providi 1982, 73-77), indirectly in the famous painted stele, in secondary use, also from Mycenae (Immerwahr 1990a, 194 My no. 21), as well as in the reddish spots on the floor of the little shrine 117 in the lower acropolis of Tiryns (Boulotis 1988, 37).

TRAVELLING AND DIFFUSION: GENERAL REMARKS

The concept of travelling, visiting or itinerant painters has always been admitted (sensibly, though usually tacitly) to explain the transmission and diffusion of mural painting within Crete and particularly outside it in the Aegean during the Late Bronze Age. By this concept I mean usually small groups of artists, dependent on a central authority but in some cases very probably also freelance, who left their home base to undertake shorter or longer journeys for the purpose of exercising their art by order, on commission or according to demand. At the same time, we should imagine that these artists did not, on the whole, undertake particularly long and hazardous sea voyages on a purely speculative basis, that is in search of opportunistic and unpredictable commissions.

The activities of travelling fresco painters are thus in broad terms easily incorporated within the more general phenomenon of the inter-regional circulation of artists and craftsmen, which is often documented or plausibly suggested as a consistent element of manufacturing and commercial transactions in this era (Keramopoullos 1930, 38-40; Kardara 1968; Branigan 1983; Dickinson 1984, 116; Poursat 1990; Cline 1995b; Bloedow 1997; Papadopoulos 1997). Invaluable hints as to this are provided by the Linear B tablets from Pylos and Knossos, some of which list craftsmen (for example, toko-do-mo = τοιχοδόμοι, 'builders'; te-ko-to-ne = τέκτονες, 'carpenters') who are obviously attached to the palace but absent at the time of recording, probably because they have been detached for service elsewhere (Ventris and Chadwick 1973, esp. 171, 174, 182; Chadwick 1979). The Odyssey (xvii.383-386) supplies us with the earliest explicit testimony on craftsmen, along with poets, healers of ills etc., as wandering professionals for hire in Aegean societies (Finley 1979, 36-37, 56; Russo et al. 1992, 38-39; Panagl 1995). According to Homer, craftsmen (δημιουργοί), including builders, are "κλητοί γε βροτών έπ' 'απείρονα γαίαν" ("on call all over the boundless earth"). In the context of Homeric society, these δημιουργοίshould be envisaged as freelance itinerants, by contrast with what seems to be the position of the majority of the δημιουργοί in the palatial societies of Crete and Mycenaean Greece. The Homeric evidence, however, is not without relevance to the issues which concern us, given that, to a large extent, the Homeric poems reflect long established industrial and commercial practices of the second millennium BC. Moreover, in that period, alongside the clearly predominant palace-controlled or royal trade (Keramopoullos 1930; Alexiou 1953-54; 1987; Kopcke 1987; Wiener 1991; Cline 1995a), one may plausibly suggest the activity of independent or semi-independent merchants (Warren 1989; 1991, esp. 251-253; Cline 1995b, 278-281); and this could also be the case for mural painters, even within the framework of the palatial systems. This would apply mainly in two sets of circumstances: a) during the advanced stages of diffusion of painting practice, in sites of non-palatial character such as the flourishing harbour towns of the Cyclades in the LC I period or in settlements far away from palatial centres; and b) in periods of upheaval or of the abolition of the central authority, as, for example, during the establishment of a Mycenaean dynasty on Crete or during possible inter-dynastic conflicts in the Mycenaean palatial centres. For the widespread prevalence of the phenomenon of travelling craftsmen, including mural painters, in the palatial societies of the era, illuminating evidence is available from the Near East, as C. Zaccagnini has shown in an article concerning patterns of mobility among craftsmen (Zaccagnini 1983; cf. also 1987, 57; Shaw 1995, 112). On the other hand, the shipwrecks at Cape Gelidonya (Bass 1967, 163ff.) and Ulu Burun (Kaş) (Bass 1986) constitute tangible evidence of the free movement not only of goods but also of craftsmen between the Levantine and Aegean worlds. Although both shipwrecks postdate the first dynamic phase of the spread of Aegean painting overseas in LM I, they provide us retrospectively with a glimpse of the probable sea routes followed by groups of Aegean painters, such as those who appear to have been at work in the sixteenth century in the Syro-Palestinian palaces of Tel Kabri and Alalakh and at the palatial complex of Tell el Dab'a in the eastern Nile delta. The possibility, however, of direct voyages from southern Crete to Egypt and vice versa, as has recently been argued convincingly by L.V. Watrous and others (Watrous 1992, 172-173, 175-178; Warren 1995, 10-11 with further bibliography), opens an alternative, considerably shorter maritime route for the Cretan painters who are assumed to have worked at Tell el Dab'a.

The inter-regional circulation of painters, who were involved (as I believe) in three basic patterns of diffusion, ⁽¹⁾ proved over time to be a dynamic and seminal factor in the history of Aegean fresco painting, which in turn had a significant impact on various other aspects of artistic creation. By 'transporting' their art from one locality to another, mural painters effectively contributed to the creation and establishment of an inter-regional painting thematic, of a common iconographic language, and (up to a point) of a uniformity in style - elements varying from period to period and displaying certain natural elements of local divergence and personal choice at a synchronic level, which L. Morgan has aptly characterised as 'idioms' (Morgan 1985, esp. 7-14). This, which amounts more or less to a koine in painting, constitutes only part of a general artistic koine, which during its first phase in the sixteenth to fifteenth centuries BC spread from Crete to the southern Aegean islands and Mycenaean Greece. Murals played an effective and important role in the gradual formation of the artistic koine, since, being exposed to common view like an open illustrated book, they affected minor arts, such as glyptic, ivory-and metalwork and especially pottery, in several ways. The frequent propinquity of workshops (in palaces and settlements) to painted spaces facilitated such influences. At the same time, the multiple activities of workshops in producing prestige items with figurative and/or decorative themes, the capacity of some artists to work in different (similar or complementary) artistic  fields, and not least the complex circulation patterns of prestige goods - even far distant from their place of manufacture - gready accelerated the spread of the artistic koine. Furthermore, in the light of the presumed activities of Aegean painters who travelled as far as the Levant and the Nile delta during the LM IA period, we can now even better understand the channels through which foreign thematic and iconographic influences passed into the Aegean repertoire of painting. The most evident among these are the Egyptian influences, such as those recognised by S.A. Immerwahr in the fresco of Room 14 in the villa at Ayia Triada (Immerwahr 1985), by eC.G. Doumas in the repertoire of the Akrotiri frescoes (Doumas 1985), and recently by S. Hiller in a comprehensive review of Cretan, Theran and some Mycenaean frescoes (Hiller 1996, with further bibliography; see also Morgan 1995, esp. 31). This channel of transfer remained open even later, with some possible Egyptian influences on Cretan and Mycenaean painting of the Late Bronze Age II and III periods (see, for example, Sapouna-Sakellaraki 1973, 300-301; Lurz 1994, 152-183; Hiller 1996, 90-91; Bietak here, vol. I). Indeed travelling painters and other artists and craftsmen, together with travelling decorated objects and presumably the use of 'pattern books', favoured an unlimited diffusion of artistic ideas (see more generally, Crowley 1989).

II was natural for earlier researchers to attribute the emergence of mural painting outside Crete, including specific compositions considered to be the earliest from the sites in question, to Cretan, particularly Knossian, travelling painters. Thus, for example, Rodenwaldt, relying mainly on thematic and stylistic criteria, had no doubt that "die ältesten Malereien in Tiryns, Mykenae und Orchomenos werden also wohl von kretischen Malern ausgeführt sein, die dort Werkstätten begründeten" (Rodenwaldt 1912, 202). Nevertheless, although recognising a close relationship between the murals of Tiryns and Knossos, he never considered the evidence sufficient to allow identification of the hand of the same painter at these two sites. Evans, by contrast, believed that by comparing composite decorative zones from these two palaces he could reasonably attribute them to the same painter; and on the basis of these particular murals he referred for the first time, in explicit albeit tentative terms, to the phenomenon of itinerant painters: "May we, perhaps, conclude that there were at this time itinerant artisans belonging to the same guild of decorators who executed works indifferendy for the Cretan and Mainland lords?" (Evans 1935, 876; cf. also Evans 1930, 304-330; 1935, 786). Similarly, Keramopoullos and Glotz attributed the procession of life size women from the Old Kadmeion at Thebes to Cretan painters (Keramopoullos 1930, 33, 40; Glotz 1953, 58ff.), while Reusch believed more specifically that the outstanding master responsible for this fresco originated in Knossos (Reusch 1956, 46; for a more moderate version, cf. Rodenwaldt 1919, 99). For his part, Evans also recognised in the flying fish fresco and the fresco with small female figures from Phylakopi the work of Knossian artists executed on the spot (Evans 1921, 542, 544, 546), thus diverging on this point from Bosanquet's unjustified view that the flying fish fresco was exported ready made from Knossos, enclosed in a wooden frame (Atkinson et al. 1904, 71). The fragments of murals of Cretan character discovered sporadically during the early excavations at Miletus (Weickert 1957, 109-110, fig. 4) and at Ialysos/Trianda on Rhodes (Monaco 1941, 91-183, pls. VII, IX, XI) were also inevitably viewed in the same light.

However, the diffusion of the art of fresco painting, which was considered either self-evident or obscure (and in most cases both), did not preoccupy earlier scholars, except en passant and in the most general manner. The socio-political parameters of the subject, let alone its wider economic aspects, lay mainly beyond the limits of consideration. The same has generally also remained true of modern research, that is until the recent discovery of frescoes of Aegean character in the palatial complexes at Tel Kabri in Israel and Tell el Dab'a in Egypt, first published respectively by Niemeier (1991) and Bietak (1994; 1995). Even in S.A. Immerwahr's fundamental monograph of 1990 on Aegean painting, one finds only sporadic allusions to the inter-regional circulation of painters. For instance, she considers the later frescoes from Ayia Triada, along with the closely related painted sarcophagus, as probable works of "emigré artists following the destruction of the palace [of Knossos]" (Immerwahr 1990a, 3). At the same time, she visualises "Minoan artists who had migrated from their homeland" in order to explain the emergence of mural painting at Akrotiri, "itinerant artists from Knossos, or possibly from Thera" at work at Ialysos/Trianda, "travelling artists" moving between the Mycenaean palaces, and "a transfer of some artists" from Crete to the mainland at the time of the destruction of the Knossos palace (Immerwahr 1990a, 4). In the case of Crete, Cameron (1975), having distinguished the characteristics of isolated workshops (palatial as well as regional), and having recognised the hands of some individual painters, occasionally tackles the problem of the movement of commissioned Knossian painters on the island, and their activities beyond the palace and its immediate surroundings (Cameron 1978, 588). More specifically, he attributed the fresco with the cult scene in a floral landscape from Room 14 of the villa at Ayia Triada to the workshop which decorated the Caravanserai or the House of the Frescoes at Knossos (Cameron 1975, 360, 466, 569-570; cf. Militello 1998, 274-275). In his opinion, the cypress trees from House A at Prasa must have been painted by the same group of Knossian painters who also executed the murals in the House of the Frescoes and the South House (both in the immediate vicinity of the palace) as well as the fresco from Savakis's Bothros in a northern suburb of the town (Cameron 1976, 7-8, 11-12 with n.20).

As for the Cyclades, the question of the transmission of fresco painting from Crete, and its diffusion among the islands through the circulation of painters, is usually examined only briefly as a minor aspect of more general considerations. Scholars have mainly concentrated on demonstrating the particularity and artistic autonomy of local workshops in relation to Crete (see especially Davis 1990; Morgan 1990; cf. also Cameron 1978; Sapouna-Sakellaraki 1981) and on distinguishing the hands of individual painters (see especially Televantou 1992a; 1992b; 1994, 300-307, 375-383). Most recently,M.C. Shaw (1996), operating within more or less the same framework, has re-examined and reconstructed in an exemplary way the fragments of the bull leaping fresco discovered below the Ramp House at Mycenae. These are of fundamental importance to the question of the transmission of mural painting to the mainland, on the one hand because of their relatively early date by Mycenaean standards (LH IIB-LH IIIA1) and on the other because of their close links with Knossian models.

The recent discovery, already mentioned, of fresco paintings in palatial buildings at Tell el Dab'a and Tel Kabri and, to some extent, the re-examination of fresco material from Yarim Lim's palace at Alalakh and the palace at Qatna in Syria, which also appear to have Aegean affinities (Niemeier and Niemeier this volume), have effectively rekindled discussion of the subject of the transmission of this art form overseas by travelling Aegean painters (Niemeier 1991; Bietak 1994; 1995; Shaw 1995; Morgan 1995). Apart from the new perspectives which they open up on the issue of the eastern Mediterranean koine, of which the Aegean world might be said to form the westernmost member, these frescoes provide us with an exemplary case study for our topic because of the great geographical distance involved, combined with the alien cultural environment in which they appear. At the same time, they stimulate us to reconsider, within a new spectrum, the diffusion patterns of fresco painting and the mechanisms which spread them within the Aegean and beyond.

Apart from hypothesising in general terms, is it really possible to distinguish precise directions of transmission (in other words, the patterns of diffusion of the art of fresco painting) on the basis of archaeological data and related evidence of a socio-historical nature? I believe that this question can satisfactorily be answered in part by addressing the following criteria: a) the geographical and quantitative distribution of the total fresco material known to date;⁽²⁾ b) the clarification of matters of relative chronology - which, however, is not always possible, as a result of stratigraphical uncertainties, particularly in the case of earlier excavations of palaces (such as Knossos or the larger mainland palaces) where mural decoration was repeatedly renewed (Hawke Smith 1976; Boulotis 1992, 82-87; Hood, here, vol. I); c) systematic study of the inter-regional convergences and divergences in theme, iconography, style, technique and quality, as well as the distinction of painting workshops and possibly of the hands of specific painters (Rodenwaldt 1912; Morgan 1990; Davis 1990; Cameron 1975; Televantou 1992a; 1992b; Shaw 1997); d) such historical and socio-political parameters as are discernible as having stimulated and propagated the expanding 'fashion' in fresco painting, which may have varied from case to case.

This question of 'fashion' in the Aegean context should be approached first of all in relation to the tendency to acculturation and the general principle of the imitation engendered by an 'ideal, brilliant model' - a process which M. Wiener has aptly compared to the 'Versailles effect' in characterising the diffusion of Minoan material and symbolic elements (including fresco painting) in the south Aegean, particularly in the Cyclades (Wiener 1984; 1990; cf. Melas 1991, esp. 173-174, 180-182, 188). Following the logic of the 'Versailles effect', and with the palace of Knossos as the dominant model, the practice of fresco painting as a prestige indicator spread to minor palaces, villas and towns throughout Crete. Among all the island settlements of this era, Akrotiri offers, in this respect, an example par excellence of the strong orientation of its elite society towards the Knossian model, with regard not only to its outstanding frescoed mansions but also to the palace itself (Cameron 1978, 589; Boulotis 1992, 89-90). On the other hand, the great Mycenaean palaces of the mainland, which had the LM II-IIIA1 palace at Knossos as their ideal model, in turn provided models for the painted decoration of smaller palatial installations (e.g. Gla, Midea) and of important houses outside the citadels (e.g. the House of the Oil Merchant and the Plakes House at Mycenae, and the LH IIIA frescoed building at Argos) (Boulotis 1995, 18-19).

In particular, as far as the role of Knossos as a dynamic originating centre for the diffusion of fresco art is concerned, it is evident that murals and painted reliefs not only had a decorative and prestige function, but also emphasised in a symbolic manner what archaeological data prove and later myths reflect: the supreme significance of the palace not only as a political, economic and artistic centre, but also as the pre-eminent religious focus of the neopalatial era (Boulotis 1987; Cameron 1987; Davis 1987; Hägg 1987; Marinatos 1987; Niemeier 1987). Indeed, symbolic themes, like bulls and bull leaping, for instance, seem to function as emblems of the palace itself (Betts 1967, esp. 22, 26-27; Shaw 1995, 98-104, 113-119; 1996, 190). In a period of intense ritualisation, such as MM IIIB-LM I, the spread of specific religious-symbolic painted themes from Knossos to various localities should not simply be explained by the desire of the elite, in Crete and elsewhere, to imitate the Knossian model according to the 'Versailles effect', but at least in certain cases should also be considered as a pictorial expression of close palatial connections or dependence - and even as a form of latent palacial propaganda. If so, we are not far from N. Marinatos's definition of 'threskeiocracy', as applied to the diffusion of Minoan religious elements, including murals, in the Cyclades in general, and specifically at Akrotiri (Marinatos 1984). Traditionally, religion constitutes one of the most effective channels for the exercise of authority; and, within this framework, widely recognised images may convey, transmit and consolidate ideology.⁽³⁾ The neopalatial rulers of Knossos, to whom some of the smaller palaces and villas throughout Crete belonged "in the fullest sense", as M.S.F. Hood argues (Hood 1983, 132; cf. La Rosa 1995), are likely to have been interested in controlling which painted themes were deployed and where within their realm of jurisdiction (cf. Cameron 1975, 271; 1978, 579; Militello 1992, 111-112; Rehak 1997). Similar mechanisms must have led to the sending of commissioned painters overseas by the Knossos palace. Nevertheless, it is probable that the specific political-symbolic motivation behind such expeditions will have differed according to the occasion. The dynamic diffusion of Knossian fresco art in southern Aegean settlements - the explanation of which depends, to a large extent, on our acceptance of the historical content of the legends concerning the thalassocracy of Minos (whether it be a question of colonies, political domination, emporia or acculturation of local elites) - arguably has to be understood in a different light than do the assumed activities of travelling painters in the earliest Mycenaean palaces, and especially in palatial contexts within quite alien cultural environments, such as Egypt (Tell el Dab'a) and Syro-Palestine (especially Tel Kabri). By far the most convincing explanation for such distant expeditions by artists, which required planning and, quite certainly, an agreement of some kind between the administrations concerned, would be that of diplomatic relations. Egyptian and Near Eastern records, combined with the archaeological and pictorial evidence, clearly testify that, during the Late Bronze Age, rulers of distant regions exchanged precious gifts and raw materials as a form of royal commerce (Keramopoullos 1930, 38-40; Zaccagnini 1973), and Cretan rulers also appear to have been involved in this from the mature Old Palace period onwards (see, for example, Boulotis 1987, 155; Wiener 1991, 328; Niemeier 1991, 199-200; cf. Niemeier and Niemeier this volume; Cline 1995a; Warren 1995, esp. 11-14). Within such inter-palatial exchange, the dispatch of skilled craftsmen and artists, acting as commissioned 'state workers', took place intermittently (Zaccagnini 1983, 250-254; Cline 1995b, 278-279). In this context, the Ugaritic myth concerning Kothar wa-Khasis, the god of handicrafts, which is preserved in the Anat epic, acquires particular significance, in that it exemplifies real practices which even have reference to Crete itself: the god has to be brought from his throne in Kaphtor (kptr -almost unanimously identified with Crete) in order to build a palace for the god Baal and furnish it luxuriously (Kapelrud 1952, 85-86; Gordon 1966, 22-23, 48-49; Ginsberg 1969, 133-134, 138; Kardara 1968, 98; Niemeier 1991, 199; Niemeier and Niemeier this volume). Along the same lines, we should bear in mind another possibility which could have led to the long-distance activities of Aegean painters in foreign lands: namely, the Egyptian and Near Eastern practice of arranging dynastic marriages for diplomatic purposes (Schulman 1979; Crowley 1989, 265-266; Liverani 1990, 274-282; Warren 1995, 11-12; Niemeier and Niemeier this volume). "It is not implausible", notes E.H. Cline, "that Aegean royalty found their way into Egyptian and Near Eastern palatial courts" (Cline 1995b, 277-278). Such an assumption would make the activities of Aegean painters at Tell el Dab'a seem more comprehensible, as Bietak has argued (Bietak 1992, 28; 1994, 58; 1995, 26; 1996, 80-81). At least in theory, dynastic or elite intermarriages within the Aegean world (especially between Mycenaean palatial centres, as reflected in myths of apparent Mycenaean origin) may also have provided an additional stimulus for the inter-regional mobility of painters, builders and other craftsmen.

MOBILITY, RESIDENCE AND ALTERNATIVE OCCUPATIONS FOR PAINTERS

We should not by any means imagine all painters on the move, nor should we consider those who transported their art around as full-time itinerants. Inter-regional mobility constituted the alternative to permanent settlement and activity in a single locality. Such complementarity within classes of painters appears even more significant in view of the following observations:

     a)    Many of the itinerants - and particularly those of the first generation - may have been trained in resident palatial workshops, or may have been in contact with them both during the time of the radiation from Knossos of mural painting and during its diffusion through the Mycenaean mainland in LH IIIA-B, with the great palatial centres again acting as initial centres in the latter case.

     b)    Itinerant painters must themselves have constituted a nucleus for the creation of new, local workshops (as full-or part-time residents), such as those of the Cyclades and the mainland.  

     c)    Out of those newly established workshops new itinerants must have sprung, and so on in a ramified process of chain reactions. Thus, for a more global appreciation of the mobility of painters it is also necessary to take into account the presence of workshops in the great centres.

The extensive, planned and periodically renewed fresco decoration of palatial centres, such as those of Knossos, and of Mycenae, Tiryns, Thebes and Pylos on the mainland, undoubtedly imply the permanent existence of workshops, whose varied activities were arguably closely dependent on the central authority. Particularly noteworthy, according to this view, are the highly qualified specialists who created the time-consuming painted stucco reliefs which were almost exclusively restricted to the palace of Knossos (Kaiser 1976).

Judging from the centralised nature of the palatial machine, apprehended only indirectly in the case of pre-Mycenaean Knossos but later directly through the Linear B evidence from Crete as well as from the mainland, it seems reasonable to believe that the central authority would effectively have overseen the establishment and organisation of the painters' workshops, their orderly operation and their provision with essential requirements such as pigments.⁽⁴⁾ Under such an arrangement, the painters' activities would be determined mainly by the central authority, which would organise expeditions away from the palace (including perhaps overseas ones), especially for its own material, political or ideological benefit. As such, the palace would not only regulate the circulation of its painters, but would maintain and co-ordinate an uninterrupted continuity of artistic tradition by training apprentices in the various palatial workshops, in the case of Knossos defined by Cameron (1975, 269, 306-373) as "schools". In this connection, I would like to emphasise the Linear B evidence for masters and apprentices in a palatial context, clearly attested in the Ak series at Knossos by the words di-da-ka-re (διδασκάλει, 'at or in school') and de-di-ku-ja (δεδιδαχυίαι, 'having served their apprenticeship, or having been trained') and by their acrophonic abbreviations di and de (Tzachili 1997, 272-280 with references). Although this evidence concerns the art of weaving, I nevertheless consider it indicative of the care taken generally by the palace to institute and control 'schools' of craftsmanship. The art of mural painting in particular, which by its nature requires collaboration on a single project, suggests the existence of a hierarchy among the members of a workshop, consisting of an outstanding master and several more or less skilled assistants who would naturally begin their careers as apprentices. Comparative analysis of contemporary Knossian frescoes, particularly those of the miniature class, combines with the continuity in general style observable from one period to the next to indicate the existence of 'master-pupil' relationships (Cameron 1975, 269, 319ff.).

Relationships of this sort, in my opinion, become even more apparent in the frescoes of the mainland. From the beginning of LH IIIA onwards, we can observe, on both a synchronous and diachronic basis as well as at different localities (especially in the case of the Argolid and Boeotia), not only an insistence on stereotyped pictorial and decorative themes (figure-of-eight shields, processions of women, hunting and battle scenes, running spirals with papyrus flowers etc.) but also an attachment to a more or less standardised iconography and choice of colours. To these can be added a notable conservatism of style, which evolved only extremely slowly over several generations of painters. This repetition, standardisation and stylistic conservatism of Mycenaean frescoes - which in these respects can be compared with the main characteristics of Egyptian painting - reveals a strict continuity of artistic tradition which remained virtually unaltered as a result of the system of training in palatial workshops and must have spread geographically by means of travelling painters, probably using some sort of pattern books. At the same time, the clear thematic, iconographic and stylistic dependence of a substantial number of earlier Mycenaean murals on their Knossian counterparts (in the case of bull leaping, figure-of-eight shields, composite decorative bands etc.), despite minor divergences in detail (Rodenwaldt 1912; Shaw 1997), provides a vital clue to an even earlier phase of apprenticeship in the Knossian palatial workshops, which gave birth to the painters who became active in Mycenaean Greece.

It is reasonable to assume further that the non-palatial workshops, which in most cases probably operated on a freelance basis, must also have been organised on the palatial model, in other words on 'master-apprentice' relationships. Moreover, it seems likely that the guild structure, based essentially on parental lineage, may have been tighter than was assumed by Cameron (1975, 269) to be the case for the Cretan palatial workshops.

Thanks to its abundant murals and their preservation in their original architectural setting, Akrotiri provides us with illuminating data concerning the activities of groups of painters, including outstanding masters and less skilled assistants, who worked simultaneously on the decorative programme of a house, and perhaps even on the same composition (cf. the West House, Xeste 3 and Xeste 4). In contrast to Knossos and the Mycenaean palaces, however, at Akrotiri we are deprived of the opportunity to follow the notion of apprenticeship in a diachronic dimension, since all the hands which have been distinguished belong to the same, post-seismic chronological horizon. This is also the case at most of the remaining sites, with the probable exception of Ialysos on Rhodes where Casa 1 has supplied us with murals belonging to three successive horizons (Monaco 1941). At Akrotiri, the volcanic destruction abruptly interrupted the local painting tradition; and, although there are indications of the existence (at least on a restricted scale) of mural decorations in the immediately preceding phases (Doumas 1992, 146, figs. 149-150; Televantou 1992a, 65 with n.25), these are too sparse to permit the direct establishment of the missing link with the Cretan 'schools' of painting.

Returning to the issue of the parallel, complementary activities of travelling and resident painters, I would like to pose the following question: Were the elite classes of flourishing settlements able to support permanently active workshops,of the sort that we assume existed in the palatial centres? The answer to this question will affect our speculations, since it touches on one of the basic reasons for the mobility of freelance or semi-independent painters - namely, that lack of employment in one location of necessity leads highly qualified specialiststo search for it elsewhere or to adopt alternative activities, or both.

The extent of mural decorations, their periodic renewal and, to a lesser degree, their quality constitute in combination the best evidence for the existence of resident workshops. This has been argued for the post-seismic phase of Akrotiri (Doumas 1983, 123-124; Michailidou 1995, 176-177; Morgan 1990) and seems generally convincing. Nevertheless, I believe that circumstances would have forced some Theran painters to take their skills elsewhere.

The nature of fresco painting, with its inherent requirement for rapidity and the simultaneous cooperation of many artists on a single project, which is well attested not only in Crete (Militello 1998, 273-274) and at Akrotiri (Davis this volume) but also in Mycenaean Greece (Rodenwaldt 1912, 109; Boulotis forthcoming), also implies the rapid completion of the decoration as a whole. This, in any case, would have been required for obvious practical reasons, since only when the decorations were complete could the space be used or occupied; and it perhaps offers the best explanation for the simultaneous employment of at least two highly skilled masters together with their assistants in the most extensive and ambitious decorative programme of Xeste 3. Under these circumstances, all of Akrotiri's murals discovered so far could have been completed in little more than a few months in terms of labour-hours - a comparatively short time in relation to the career of any individual painter.

What, then, did the painters of Akrotiri do while awaiting the next uncertain order for mural decoration, particularly since this was hardly a routine feature in the regular life of a building? Part of the answer depends on the function of two uncertain factors: the number of houses which may have contained mural decorations within the unexcavated area of the settlement, and the number of painters or workshops which may have been in operation simultaneously. As it is, evidence available so far indicates that there must have been a large number of painters involved during this particularly enthusiastic phase of mural painting. Against this, however, and despite the widespread 'fashion' for murals, there were almost certainly socio-economic constraints in operation, as in all hierarchical societies. I, for one, therefore find it difficult to accept that the elite of Akrotiri would have been able to support more workshops than are already known, or even to maintain these on a full-time, permanent basis. As a result, it is possible that some painters may have departed to look for work elsewhere as itinerants, while others may have remained to combine part-time or periodic mural painting with other related activities.

One area of closely related work was almost certainly the manufacture of stucco offering tables of the sort found within the settlement, either plain or painted with figurative or decorative themes in fresco technique (Doumas 1992, 180, figs. 142-144). Their inter-regional spread in this and the immediately succeeding periods (Polychronakou-Sgouritsa 1982), together with the great demand for them, would have consumed a portion of the mural artists' labour time. It indeed seems probable that offering tables were mass-produced in painters' workshops, as attested by almost fifty specimens stored in the villa at Nirou Chani (Xanthoudides 1922, 15-16), and that the decorated ones in particular constituted a prestigious export product. In this connection, it is worth mentioning the fragments of an offering table with a miniature frieze of blue birds and vegetation in a rocky landscape, painted against a white background, which were recently found in the foundations of the palace at Tiryns, with a consequent terminus ante quem of LH IIIA1. The southern Aegean origin of this example is evident in many of its aspects, with its closest parallels dating from the LM I/LC I period; and we should not exclude the possibility that it may have been imported from Akrotiri itself, or made by some itinerant Theran painter or one of his apprenticesat a later date (Boulotis 1992, 85, pl. 37a). At least in theory, mural painters could also have found occasional complementary employment in the field of quality pottery, by being commissioned to execute pictorial decoration, which was evidently highly appreciated. The close interactive links between vase painting and murals, beginning with Kamares pottery, have been documented in detail by several scholars (cf. e.g. Evans 1930, 308-313, 381; Hood 1978, 47 with references). As I have suggested, following the similar views of Cameron and Walberg, it is likely that the first Minoan mural painters of the Old Palace period actually sprang out of the Kamares potrery workshops (Boulotis 1993, 15-17; 1995, 14-16). Close thematic and stylistic connections between mural and ceramic motifs have repeatedly been suggested for Akrotiri in particular (especially Marthari 1987; 1992; Immerwahr 1990b; Davis 1990, 223; Boulotis 1992, 223). Thus, the occasional movement of painters from one art form to another appears entirely plausible, with pottery decoration providing a potentially steady source of alternative employment. Such differences as can be observed between Theran fresco motifs and related ceramic ones can probably be put down to the differences in artistic media, and perhaps to the likelihood that it was the less skilled assistants, rather than the masters who devoted themselves to vase painting. Some of the outstanding masters must, I believe, have numbered among those painters forced to travel abroad because of lack of orders at home, if only because it is difficult to conceive that highly qualified specialists, like the charismatic miniaturist of the West House, could have involved themselves for long in activities inferior to their art.

If the above hypotheses hold good at Akrotiri, they are likely to apply even more strongly to settlements where mural painting never became a widespread 'fashion', but was used strictly selectively to decorate only those architectural spaces with specific functions or a few elite buildings. I am more or less convinced that the limited pictorial decoration found at many individual sites throughout the entire history of Aegean fresco painting is generally due to the work of travelling artists, who were either dependent on a central authority or worked freelance. From this perspective, the murals of certain Minoan villas (for example, Ayia Triada, Tylissos, Amnissos, Prasa) or of settlements like Pseira (with its stucco reliefs) constitute exemplary cases of the work of visiting artists, apparently of Knossian provenance. In the case of the frescoes from Palaikastro, Cameron assumed the existence of a local workshop on the basis of their provincial nature in relation to Knossos (Cameron and Jones 1976, 18-19). However, here again the relative scarcity of material, in my view, counts against the existence of a permanent body of full-time painters, and instead it is perhaps more reasonable to posit the presence of itinerants, probably based at Palaikastro and active within the wider region of eastern Crete. We are led to similar conclusions from a study of the small numbers of pictorial frescoes at Mycenaean sites (especially in the Argolid and Boeotia) which lie beyond the palaces and their immediate urban surroundings. Returning to the Cyclades, at Phylakopi, where only the so-called pillar crypt was decorated with pictorial compositions (Atkinson et al. 1904, 70-79; Morgan 1990), there is, in my view, no justification for assuming the existence of a resident local workshop of full-or even part-time painters. This view is reinforced by the high quality of the murals, which is suggestive of experienced artists who would have been most unlikely to exercise their art solely on an occasional basis. For these, travelling must have been a necessity. Similar conclusions are also dictated by the number and distribution of frescoes in the settlement at Ayia Irini, where they are found only in a few buildings, particularly House A and the north-east bastion (Abramovitz 1973; 1980).

In view of the above, I therefore find it difficult to agree with Morgan, who, on comparing the mural material from Thera, Melos and Kea, arrives at the conclusion that there was "a common tradition with different artists working on each island". As she herself admits, "this is a surprising conclusion, considering the degree of specialisation usually associated with mural painting. One would have expected travelling artists rather than local painters at each site" (Morgan 1990, 263).

Given the flourishing nature of mural painting at Akrotiri on the one hand and, on the other, its occasional and restricted use at Phylakopi, Ayia Irini and on other islands, the obvious question arises: Could it be that Akrotiri, after receiving the art of mural painting from Knossos, rose to become a 'metropolis' of painting, whence many of the part-time itinerant painters journeyed forth to fulfil commissions? If so, it would have played, at least within an inter-Cycladic framework, a similar role to that of Knossos on Crete and to that of the Mycenaean palaces on the mainland after the beginning of LH IIIA. Could it also be that these painters travelled as far as the Mycenaean mainland before the construction of the first palaces, as has been argued, and perhaps even further still - as far as Tel Kabri?

(1).      These patterns are: a) rectilinear radiation, b) the billiard ball effect, and c) the mixed pattern. These will be explained and discussed in detail in the larger study now in preparation.

(2).      This is, for instance, what F. Blakolmer and S. Hiller have undertaken to document in their forthcoming corpus (Blakolmer 1995, 464).

(3).      For the symbolic use of Mycenaean iconography, see also Laffineur 1992 with further references.

(4).      Evans reported several deposits of pigments in the palace at Knossos, while a large piece of imported Egyptian Blue was discovered on the acropolis at Tiryns.

----------------------------------------------------------