Christopher Chippindale

This is one of the web files where I keep published or unpublished writing, so they are available in electronic form. The purpose of putting this on the Internet is so that people can have access to it. Any fair use of it is welcome, including its use as much as you want in learning and teaching, and printing-out or using items from it in reasonable quantities. It is not fair to pass it off as your own work, or to use material in a way that generates income for you.

I am always glad to hear from people who find anything in my work useful, or interesting. Feedback is most welcome.

The texts are taken from whatever is to hand. For published work, that is usually the word-processor file originally submitted to the editor of the publication. It will be essentially what was published, but without revisions, corrections and proof-reading fixes: for a word- and letter-perfect version you must look at the publication. They also lack, usually, all the pictures, for which again you must look at the publication.

The numerous footnotes in the

To send an e-mail to me, click here, to return to my home-page click here.

Christopher Chippindale currently at

Cambridge University Museum of Archaeology & Anthropology, Downing Street, Cambridge CB2 3DZ, England
cc43@cam.ac.uk

This page is written by Christopher Chippindale. Last update: 12/2/2000. © Christopher Chippindale and/or its other author(s).

Material and intellectual consequences of esteem for Cycladic figures

David Gill
Department of Classics and Ancient History, University College of Swansea, Singleton Park, Swansea SA2 8PP
D.W.J.Gill@swansea.ac.uk

&

Christopher Chippindale
Cambridge University Museum of Archaeology & Anthropology, Downing Street, Cambridge CB2 3DZ, England
cc43@cam.ac.uk

Gill, David & Christopher Chippindale. 1993. Material and intellectual consequences of esteem for Cycladic figures, American Journal of Archaeology 97(3): 602—673.

Contents

Introduction: connoisseurship, archaeology and Cycladic figures

The material, the intellectual and the market discovery of Cycladic figures

The known corpus and its provenance

Casual finds

Archaeological investigation

'Surfacing'

Illicit excavation

New manufacture

Re-emergence of forgotten collections: a convenient fiction?

Wish-fulfilment

Material consequences

Impact on the archaeological context of Cycladic figures

Consequences for chronology

'Workshops' and regional variation

Size and function in the figures

Intellectual consequences

Reduction and corruption of the factual record: towards an impossibility of understanding Cycladic prehistory?

Art, eternal value and the idea of redemption

Identifying the masters of Cycladic art: the available information

Identifying the individual craftsman in prehistory

Identifying the masters of Cycladic art: premisses and consequences of the method

Identifying canonic proportions in Cycladic figures

Birth of the genius that was Greece

Conceptions of Cycladic and historical scholarship: the case of the baitylos

Presentations and perceptions in Cycladic studies: the choice of words and the values they carry

Cycladic figures: lying or standing?

Cycladic figures: white or coloured?

Modern perceptions of ancient Cycladic society

Art, aesthetics and human universals

Connoisseurship and archaeology: an essential asymmetry?

List of tables

Table 1. Cycladic figures from legitimate archaeological excavations.

Table 2. Security of provenances for items in Early Cycladic Art in North American Collections.

Table 3. Date of acquisition of 'Keros' items in Early Cycladic Art in North American Collections

Table 4. North American Cycladic exhibition, dates of surfacing.

Table 5. Declared provenances for items in the Karlsruhe exhibition, 1976.

Table 6. Period of acquisition of Cycladic figures into some major museums.

Table 7. Cycladic figures of male and special types.

Table 8. Figures over 70 cm long.

Table 9. 'Drifting provenances' between the exhibitions at Karlsruhe 1976 and at Richmond 1987.

Table 10. Numbers of graves and numbers of figures on some Cycladic islands.

Table 11. Declared histories of items in the exhibition, Art of the Cyclades in the this millennium BC, 22 April – 1 May 1989, Galerie Heidi Vollmoeller.

Table 12. The variation of figure size with period.

Table 13.. Security of provenance of objects in the 1979/80 loan exhibition, Greek art of the Aegean islands.

Table 14. The security of the corpus of Archaic sculpture, statues of kouroi and of korai and gravestones.

Table 15.. Security of provenance of items in the White and Levy collection.

Table 16. The security of corpuses attributed to sixteen Cycladic masters.

List of figures

Figure 1. Proportion and 'canonical' Dokathismata figures.

Figure 2. Proportion and 'canonical' Chalandriani figures.

Figure 3. Construction grids and a 'canonical' harpist figure.

Introduction: connoisseurship, archaeology and Cycladic figures

An interest in ancient things has been, and is, directed by two modern concerns. Connoisseurship we define as esteem for, and appreciation of, beautiful artefacts, especially those that seem to fall into the domain of the fine and decorative arts. Archaeology we define as the study of ancient societies by means of their surviving material remains. The two concerns overlap when it comes to those material remains of ancient societies that are regarded as beautiful.

In the field of classical studies, connoisseurship has the longer history, going back to the regard for Hellenic arts that was the motive for Roman collection, removal and copying of Greek decorative arts. For the most part the two concerns have gone harmoniously together in the hybrid discipline of classical art history (so often effectively equated with classical archaeology), which has been central to classical learning over the last centuries. But they are distinct studies, and their distinct concerns and priorities have distinct consequences. Many elements of classical archaeology have no part in connoisseurship – the rubble masonry, fragmentary coarse ceramics, and animal bones that make up the greater bulk of surviving material remains from classical antiquity. Equally, elements of connoisseurship provide no information about ancient societies: the aesthetic regard held since the Renaissance for the surface qualities of bare white marble is a recent pleasure, quite separate from the ancient appearance of Hellenic statues that were actually decorated in bright polychrome paint.

The carved marble figures from the Cycladic archipelago are instructive as to the differing concerns of connoisseurship and of archaeology. They were no part of the Greek canon, as that provided an ideal from the antique through the Roman period, the Renaissance and into the19th century. They first came to light when the Hellenic ideal still dominated the western aesthetic of sculpture, so they were dismissed as unlovely curiosities. Within a few decades, the course of western art chanced to take a path towards aesthetic forms that mimicked elements in the Cycladic figures. Accordingly, the 20th century, and especially the years since 1945, have seen the Cycladic figures, while remaining an important aspect to prehistoric archaeology in the Greek archipelago, move decisively into the connoisseur's domain. Cycladic archaeological research in those decades has been notable for aspects of excavation and field-survey of a considered thoroughness, in studying the special place of the Cycladic sequence in the larger development of Aegean cultures, and in concern for the maritime aspect of the archipelago and its settlement. The connoisseurship of Cycladic figures has seen grand exhibitions with large catalogues, a collecting boom, a substantial commercial market in figures, a spirit of competition between museums anxious to possess this new character of masterpiece, new manners of regarding the objects on the model of renaissance art, and a new museum in Athens devoted to Cycladic figures. Indeed collections of ancient sculpture, whether in private collections or in museums, are no longer to be thought intellectually and historically complete if they lack Cycladic figures. Cycladic figures are beginning to acquire names, like 'Cycladic Statue of a Reclining Woman', to fit their standing as modern masterpieces.

The same period, especially the decade of the 1960s, saw illicit looting of cemeteries and other places on the islands where figures and other fine artefacts may be found. In this, Cycladic archaeology has suffered in the manner of other areas in the classical world, notoriously the tombs of Etruscan Italy; in turn it is a prototype for what is now inflicted on, among many instances across the world, sites yielding terracotta figures in Mali, west Africa.

For these reasons, the prehistoric figures of the Cyclades offer a case-study in the material and intellectual consequences of aesthetic esteem for a particular class of antiquities. For many classes of classical antiquities, whether sculpture, painted pottery or bronze castings, the aesthetic interest has long run alongside the archaeological; for coarse-ware and farm implements, the interest still remains archaeological alone. Cycladic is unusual in that a long-standing archaeological concern has recently been joined by a new respect for its aesthetics, which has moved it decisively from an archaeological backwater into the aesthetic mainstream. This accident of history allows the two traditions, so often melded in classical archaeology, to show some of their distinctions.

Some of what is set out here is well known. We excuse the necessary length of the paper because we think the whole picture, or as much as a long paper will hold, amounts to very much more than the obvious pattern of the few fragments that are common knowledge. We have occasionally gone beyond the narrow Cycladic story when it is relevant to some more general issue; to keep the discussion in bounds, we have generally taken examples or quotations for those larger issues from a small number of representative sources. One of those larger issues is the ethical concern that may arise for museums and scholars, which has often been aired. We do not explore that, but simply document and to describe the various consequences of esteem for Cycladic figures.

Much of the recent publications on Cycladica, and the new regard for its aesthetics, is due to a small number of scholars who have specialized in the subject, especially Getz-Preziosi, Renfrew and Thimme. Our paper therefore depends on, and makes frequent reference, to their work.

The material, the intellectual and the market discovery of Cycladic figures

Cycladic figures have no part in the common canon of ancient treasures in the era of the grand tour, although reference to them can be found at an early date. Italian travellers refer to idoli and idoletti in 1771;: an early scholarly reference to a Cycladic figure dates to 1818, a figure discovered by the Earl of Aberdeen in a grave near Athens. Two formed part of the collection, formed in Greece, of the 6th Viscount Strangford (1783–1855), ambassador there, and one was left to the British Museum in 1840.

Prominent in the earliest Cycladic material to find its way to European collections are the figures that came to light a little over a century ago, excavated by James Theodore Bent during his travels in the Cyclades, 1883–4, and subsequently taken into the British Museum. The Museum built up its collection of Cycladic material in the later part of the nineteenth century; some purchases have been made subsequently. The university museums of Cambridge (Fitzwilliam Museum) and Oxford (Ashmolean Museum) seem to have derived their Cycladic collections in part through the good offices of members of the British School at Athens. The Fitzwilliam acquired its first Cycladic figures through R.C. Bosanquet in 1901 and its last in 1934. The Ashmolean started its collection in the later nineteenth century with a few items added in more recent years. The picture is somewhat similar elsewhere in Europe. The Staatliche Kunstsammlungen in Dresden had Cycladic figures by 1841, and the Badisches Landesmuseum in Karlsruhe by the early nineteenth century; Karlsruhe followed with a burst of acquisitions from the 1960s, when the Cycladic figures had gained their new and higher status.

First perceptions of Cycladic figures were hostile to their aesthetics, which is why the intellectual context of these first acquisitions was more archaeological than artistic. The Karlsruhe catalogue volume is introduced with the early opinions, 'Small monsters made of bits of marble', 'ugly', 'barbarian'. Greek and ancient, they came by degrees also to be 'art', making their entry into studies of ancient Greek sculpture in 1892 and of ancient Greek art in 1894, at first generally qualified by the word primitive; one finds the word 'primitive' again in the British Museum catalogue of 1928, which puts Cycladic figures along with some from Anatolia and one from Cyprus into a class of 'primitive idols'. Important to the growing appreciation of their artistic value were the opportunities for some great European sculptors of the twentieth century to see, and to be inspired by, the little figures. The artists' discovery of Cycladic dates to, and is part of, that turning to non-western art, modern and ancient which influenced artists of the new century. Jacob Epstein discovered Early Greek works and Cycladic sculpture displayed in the Louvre, 1902–5. The work of Constantin Brancusi so evokes the Cycladic in its simplicity of form and restriction of detail that Epstein and others believed Brancusi to have been inspired by the Cycladic figures that he could have seen in Paris. Henry Moore acknowledged affinities of his work to the Cycladic figures he saw in the British Museum from 1920, later writing of their force among the many ancient things he valued in the British Museum collections. Moore owned three Cycladic figures himself. Picasso owned one and is said to have said of it, 'Better than Brancusi. Nobody has ever made an object stripped that bare.' Barbara Hepworth's work is seen to echo Cycladic, directly or via the Cycladic look to Moore's, and Alberto Giacometti acknowledged affinities between his own work and Cycladic.

In due time these young sculptors became the mature masters of the modern movement, and with esteem for their work came esteem for the primitive and prehistoric figures which the modern masters echoed or followed. It is immaterial whether Cycladic influenced the moderns consciously or unconsciously, or whether the coincidence of forms is a mere happenstance; it suffices that the Cycladic figures fall within the same aesthetic frame. Accordingly, they are treated as belonging together, and those modern collections interested in non-western analogues to contemporary sculpture like the Sainsbury Collection at the University of East Anglia, Norwich, give Cycladic a special attention, and Cycladic has come to find a natural home in collections primarily of modern or tribal art. When the collecting of modernist sculptures in North America began after the Second World War, the collecting of Cycladica followed, well after the time when the main British collections had been formed. A great many figures surfaced in the 1960s; as Getz-Preziosi has remarked, this was in part due to the publication in 1957 of C. Zervos' study of Cycladic art.

Table 6, page 00 below, documents the dates when some major museums acquired Cycladic figures. The British collections, established early, contrast most with those in the USA, which echo the boom of the last three decades in private US collections.

In addition to the archaeological collections in Greek public museums, mention must be made of the Goulandris Collection. This important collection, after special exhibitions in Japan, the United States and Europe, has been installed in a fine new building in Athens. It has served the purpose, important for the Greek authorities, of keeping Cycladic figures in Greece, though it is made of objects which have lost their archaeological context. During the period of intense looting in the 1960s, Christos Doumas explains, 'individuals were encouraged to collect Cycladic objects and so prevent them from being take out of Greece. Thus several private collections obtained figurines and vases: the most notable of these is the Goulandris Collection which was begun at this time and subsequently enlarged to the point where it has become the largest and finest private collection of Early Cycladic art in the world'.

The Goulandris with official encouragement has kept Cycladic figures in Greece, specifically in the mainland capital of Greece – not in the Cyclades. They have been removed to the controlling centre of the modern nation-state, where they help extend the modern national self-esteem and identity. When photographed against the intense colour of the Aegean Sea, the Cycladic figures make a physical prehistory for the blue-and-white colours of the modern national flag. If one chooses to question those identities, one may doubt if Athens, Greece really is a more natural resting-place for a Cycladic figure than Athens, Georgia.

The role of the Goulandris, in gathering together 'orphaned' figures, may provide an effective floor for the market, if the market-makers believe that the Goulandris is always alert to acquire Cycladica at some agreed fair price. In so doing, the Goulandris mission may inadvertently support the market and encourage the flow of new finds into the market-place.

Permission was refused by the Greek authorities for a few important Cycladic objects to travel from 'a private collection in Athens' to the 1976 Karlsruhe exhibition, despite 'years of petitioning'. This, the exhibition organizers' 'one severe disappointment', prevented these items, held legitimately within Greece, being put on display in Germany alongside so many items whose export from Greece is not documented.

More recently, the Goulandris Collection has been a vehicle for the recovery for Greece of figurines that had left the country and then came on the open market overseas. After Court action before the Sotheby's sale of the Erlenmeyer collection in July 1990, three items were withdrawn from auction, and purchased by the Greek government; they are now in the Goulandris collection. Sotheby's went to great lengths to reassure collectors and dealers that no pieces on offer would be further pursued by the Greek government; responding to market concern, a Sotheby's spokeswoman said, 'As the Cycladic pieces were actually sold to the [Greek] government there is nothing to be nervous about.' We have noted above the role of quasi-official purchases by the Goulandris as supporting the market; these three pieces were bought, according to Sotheby's 'at prices that reflected their market value'. Others were purchased for the Goulandris from those that remained in the subsequent July auction, and from another Sotheby's sale of Cycladic items from the Erlenmeyer collection in December 1990.

Discovery by the public and private market of Cycladic figurines has matched the rise in other interests, as the trade has widened the old aesthetic range to include non-western visual arts, ancient and recent, as well as other 'collectibles', like classic cars, not previously thought of as art objects. Cycladic figures have not yet passed through public sales in sufficient numbers and for sufficient years for it to be possible to construct a reliable index of their rising values over the years, as in Reitlinger's Economics of taste, but their rising status over the decades is plain enough. One of many evident signs that the most expensive Cycladic figures have now reached a major art league is the treatment given to the star item in the Erlenmeyer sale, lot 137 at Sotheby's 9 July 1990, 'A Cycladic marble fragmentary male figure, Early Bronze Age II, said to be from Amorgos, Spedos Variety, circa 2600–2400 B.C.' No estimate is printed in the catalogue, an indication that a serious sum was expected, but 'Estimate on request'; the estimate given, when one of us requested, was £250,000–300,000, this although the piece is fragmentary and its security of provenance, 'said to be Amorgos', was further weakened by the catalogue comment remarking 'it is quite possible that it was actually recovered on Keros'.

The first stage of the aesthetic discovery of Cycladic, then, was to recognize its high place in the primitive world of 'the other': Cycladic art from the Sainsbury collection was shown at the Museum of Primitive Art, New York in 1963. A second stage is perceptible now, as Cycladic begins to fall also within the domain of the classical once that goes beyond an old and narrow definition. In its first years' purchases after its founder's endowment became available about 1981, the Getty Museum enlarged its classical holdings within the conventional Graeco-Roman range collected by J. Paul Getty himself; now it has turned towards earlier Greece, to the Archaic with a controversial purchase of a kouros of unexplained history, and to the Cycladic with the 1988 purchase of five figures from the Steiner collection. Its director John Walsh remarked in 1990 of one of these, 'There is a wonderful Cycladic figure of a harpist, now part of a distinguished group of prehistoric Aegean objects of the sort that did not interest Mr Getty at all.' Shelby White and Leon Levy began their celebrated New York collection by purchasing classical art; then their curiosity led them to wonder about the antiquities that came before and that followed ancient Greece, and took them to acquire Cycladic and Neolithic objects from Greece. And at the same time, the new esteem makes the Goulandris museum, devoted to Cycladic alone, a treasure as fine as any other in Athens. It is said that the presence of works of a single 'sculptor' such as Getz-Preziosi's 'Goulandris Master' in the 'great' museums of the world necessarily implies the figures are themselves great. As Cycladic figures now possess such virtue that they deserve fine museums, so it follows that the presence of Cycladic figures make a museum fine. And as the presence of a Cycladic Master's work in a great museum validates the standing of both the Master and the museum, so may the association of a dealer with a museum validate by its purchases the dealer's standing and judgement.

These motives and interests for the rise of Cycladic concern the aesthetics of the figures. A separate convenience arises from their size. Typically, Cycladic figures are around the length of your forearm, or rather less. In the heyday of classical collecting, the acquisition of fine antiquities went with the building of grand houses, or even whole landscapes to fit the classical manner. Now that houses are smaller, and the wealthy live more often in city apartments, finding house-room for works of art encourages the private collector to prefer the smaller object, old or new. The same pressures even affect the museums, increasingly feeling the cost per cubic metre of their galleries and fearing they will run out of space. But smaller objects, often delightful, and more practical to manage, lose by their smallness; they are less sublime, less able to impress with bulk and grandeur. Here, the Cycladic figures provide an admirable way forward. They are physically small, just the right size to stand on a Manhattan mantelpiece, yet they contrive also to be monumental in manner; well lit to make the best impression, as in the fine photographs, they look very much bigger than they really are. Where a Rothko requires a wall 5 metres square, and a Moore demands a floor reinforced to bear a couple of tonnes, the Cycladic figure is domestic in its needs. A finer view of Cycladic is part of the present privileging of the pretty portable object.

The known corpus and its provenance

Getz-Preziosi estimates the total corpus of known Cycladic figures as 1600. This number of figures has come to light, and then to public notice, in various ways:

In the term 'surfacing', we refer to items which have no known history between the ground and some remote public place. They simply appear in collections, on exhibition, or at auction, having been bought privately. When a provenance is offered, no documented account is given to connect the piece to the Cycladic island, or site within an island, from which it is declared to have come. At best the provenance must be in terms of 'said to be from' island x or y, or in Getz-Preziosi's convention given in inverted commas. These items, which come to notice without known history, may have arisen in several ways:

We have attempted to gather together, as Table 1, a list of Cycladic figures from archaeological excavations, or other find-spots of reliable security. It has proved difficult. Many figures are published or referred to in ways that leave their recent history obscure.

Casual finds

Archaeologists, on systematic survey, or walking at known sites, have occasionally come across figures, or figure fragments, as surface finds. A 'stray find' from Trypiti on Melos is reported by Zapheiropoulou. In 1963, Colin Renfrew collected four fragments from the surface of the site at Dhaskalio in Keros. For centuries, farmers must have turned up stray figures; in 1884, Bent reported of Arkesine on Amorgos, 'Ancient tools, vases, and statuettes are turned over every time [the old farmer who owns the place] ploughs.' A very few of these older finds may have been conveyed into the corpus.

Archaeological investigation

Legitimate archaeological excavation of the Cycladic cemeteries has provided many examples of figures, beginning with Bent's digs into two cemeteries on Antiparos in 1883–4. Excavations on the islands of Paros, Antiparos, Despotiko, Naxos, Amorgos and Syros have recovered approximately 143 examples from among the grave-goods in more than 1600 graves. One figure of the Cycladic type has been found on Crete in a communal tomb at Koumasa. Figures are not, however, confined to graves. Some 43 examples were found at Ayia Irini on the island of Keos, a probable thigh from a figure at Mikre Vigla on Naxos, and several at Phylakopi on Melos, in domestic or quasi-domestic contexts. One of the largest concentrations was found on Keros where 'dozens' of fragmentary folded-arm figures were found by Doumas and Zapheiropoulou, with four further fragments and one complete figure found at later dates; this has been seen as either a cemetery or a sanctuary deposit. Doumas reported 'hundreds of fragments of figurines' from this site, a number that would upturn present notions of what the Cycladic corpus amounts to. The ground at the site is littered with human bone, as well as figure fragments; whatever the site amounts to, it was in part funerary,

'Surfacing'

In addition to the excavated material, the corpus of known figurines has been swollen by the numerous examples which have 'surfaced' in the art market during, mostly, recent decades. By 'surfaced', we refer to the first appearance as a work of art of an object whose finding or excavation in the field has never been reported. These objects which appear in an exhibition or sale without history, previous publication or other account we take to have surfaced there for the first time. There are essentially three sources for this: the illicit excavation of archaeological sites; the modern creation of objects; and the re-emergence of figures from the obscurity of 'old collections' where they had been forgotten.

Illicit excavation

Since the end of the nineteenth century it has been recognized that archaeological sites were being looted to provide figures for museums and collectors. Bosanquet reported in 1897, 'Still more necessary [than museum study] is the systematic excavation of cemeteries, a branch of research which Greek archaeologists have too often left to peasants in the pay of dealers.'

Occasionally, illicit finds end up in museums when confiscated by the authorities, such as 2 figures from Naxos (Phionta). The rest go into that pool of unprovenanced and uncertainly provenanced Cycladic figures in which the private and public collectors fish.

These illicit excavations continue to dog Cycladic studies. In 1987 Paul N. Perrot, Director of the Virginia Museum of Fine Arts, noted in his Foreword to the exhibition catalogue Early Cycladic Art in North American Collections that the qualities of Cycladic sculpture 'have caused decades of clandestine looting, abuses that make it so difficult today not only to trace the evolution of the producing society but also to gain any understanding of the relationships among the various forms that have been preserved'. Getz-Preziosi, writing the same year, identified the new popularity as the specific cause of the new destruction: 'to meet the strong demand, unauthorized excavation of Cycladic sites has flourished uncontrollably and, unless strict measures are taken soon, it appears that it will end only when the sources are forever exhausted'.

In the case of the Virginia exhibition, a demonstration that the large proportion of Cycladic figures have illicit excavation as a likely source seems to be given by the declared provenances of Cycladic figures in this exhibition (Table 2). Not one is of a known provenance, rather more than a third are 'said to be', and the balance of nearly two-thirds are unknown. The Cycladic artefacts are even less securely tied to origin. In the exhibition catalogue, Getz-Preziosi looks to quite a different cause for so many unstated provenances, and says: 'very few recently acquired works are fresh finds. Many of them have come to the United States from European collections.' Whether this is an adequate explanation depends on whether they were transferred from old European collections where they had been many years, or whether they were themselves recent acquisitions into European collections; we give cause below to doubt whether old European collections can be a source of Cycladic figures in such quantity. Getz-Preziosi's statement is supported by a note, 'a majority of the objects in the exhibition are in compliance with the resolution passed by the Council of the Archaeological Institute of America on December 30, 1973'. The assurance is cheering, but no evidence is offered to support it.

The resolution, the AIA's second on plundering and looted artefacts, commended museums to co-operate in endeavours to prevent illicit traffic in cultural property by 'refusing to acquire through purchase, gift, or bequest cultural property exported subsequent to December 30, 1973 in violation of the laws obtaining in the countries of origin'. So Getz-Preziosi's reassurance, 'a majority . . . are in compliance', misses the point of the AIA resolution, which was to stop the entry of all smuggled items into museums – not to ensure that smuggled items made up only half of their acquisitions. Even in her terms, it is not an overwhelming majority. Of the 92 figures, no findspot is offered for 59. One wonders how it can be known – for this undocumented majority – that they did leave Greece before 30 December 1973, the date the AIA resolution came into force. The other 33 have a declared provenance – that is, some declared place of origin – though not a single piece comes from a secure, properly described archaeological context; there is simply a 'said to be', the usual undocumented and unsubstantiated assertion that the item in question comes from somewhere or other. Whether an item falls within or beyond the AIA pale (equally, whether within or beyond the Greek law) depends on when it left the Cyclades and Greek territory, so statements that are vague as to place and silent as to date are not material.

Some of the declared Karlsruhe provenances may not be correct. Take for example the pair of marble figures (cat. nos. 9 and 10) and the pair of marble pyxides (cat. nos. 138 and 139). These first 'surfaced' in 1971, in the gallery of an antiquities' dealer in London. The objects were displayed at Karlsruhe in 1976; appendix 5 to the Karlsruhe catalogue has a discussion of 'An early Keros-Syros grave group "from Marathon"'. By 1987 Getz-Preziosi had doubts on the security of this archaeological grouping, although she repeated the provenance of Marathon. She pointed out that although the figures belonged to the ECI horizon (c. 3000–2800 BC), the pyxides belonged to ECII (c. 2700–2200 BC). Either this was a grave used perhaps over 800 years, or the objects were not found together; if the latter, then the statement of association falls, and the accompanying provenance of Marathon is made suspect. In regard to provenances in general, Getz-Preziosi says, 'for the great majority of the known pieces of sculpture the precise find-places have been lost (or, worse, falsified)'.

A previous AIA resolution, of 1970, condemned 'the destruction of the material and historical records of the past by the plundering of archaeological sites both in the United States and abroad and by the illicit export and import of antiquities'. A celebrated illicit destruction of a site occurred on Keros in the early 1960s. More than 350 fragmentary figures are said to have been derived from this context; they were then purchased as a single lot by Professor Hans Erlenmeyer of Switzerland. Further fragments from Keros were dispersed, although in 1975 Getz-Preziosi was able to study many pieces with the 'original' owner. The passing of time and the association of the fragments with a named collection were not sufficient to prevent criticism and even outrage when the Erlenmeyer Collection came up for auction at two sales at Sotheby's in London during 1990.

The North American Cycladic exhibition demonstrates how museums and private collectors alike have ignored the loss of context when faced with the opportunity to acquire. Sixteen figures in the exhibition are said to have come from Keros. The provenance of one of these, number 73 (acquired by Fine Arts Museum of San Francisco, in 1981), is stated in the catalogue as 'Reputedly found on Keros'; this figure was published elsewhere by Getz-Preziosi as coming not just from Keros, but specifically from the Keros Hoard. The dates of acquisition of 'Keros' figures (Table 3) are as usual instructive, remembering that the Keros hoard came into the market in the early 1960s. As it becomes clear that the Keros site contains, or contained before its looting, several or many hundreds of figures and figure fragments, so it becomes clear that tracing single figures or groups of figures 'said to be' from the 'hoard' is not very material to making sense of a site which seems to be much more than the single closed group that is implicit in the word 'hoard'.

Since there was once a time when Cycladic figures, like other antiquities, could be moved more freely between national legislations, a break-down of when the figures in the exhibition first surfaced is of relevance (Table 4).

Using the 1970 AIA resolution as a watershed (as Getz-Preziosi suggests) some 47% (43 ex 92) of figures surfaced after the resolution. The majority of these are in private American collections. Their appearance in an exhibition such as this publicizes, celebrates and legitimates the pieces, to some extent making them authentic and respectable; from being just 'From a private collection', they can become 'Exhibited at Virginia Museum of Fine Arts, Kimbell Art Museum, The Fine Arts Museum of San Francisco'. This celebrity can raise their value should their owner wish to dispose of them. Attribution to named masters has the same effect.

The history of the Hunt collection illustrates this role for a high-status, high-profile exhibition in raising esteem for a collection, with a financial benefit when in time it comes to its dispersal. The Hunts' Greek vases and Greek, Roman and Etruscan bronzes were celebrated in their 1983 exhibition at the Kimbell Art Museum, Fort Worth (TX), entitled 'Wealth of the ancient world', whose catalogue was published in association with the Californian dealers Summa Galleries, from which some of the pieces had been acquired. When hard times and tax demands overwhelmed the Hunt brothers, the collection was sold in New York, June 1990, in a much-hyped sale which achieved exceptional prices. The Hunt collection did not include Cycladic pieces. The Thétis Foundation collection, which followed much the same route of public celebration before a dispersal sale, did. Exhibited and celebrated at the art and history museum, Geneva, in 1987, it was auctioned in London in May 1991; the first 22 items were Anatolian and Cycladic items. The auction catalogue indicates that they surfaced in the usual way. Of the 14 Cycladic items, the only references offered in the sale descriptions for 13 are to the catalogue of the 1987 Thétis exhibition at the Geneva museum; the 14th, an anthropomorphic beaker, was in Karlsruhe in 1976, when its provenance was declared as 'unknown'. The prices achieved for fine Cycladic figures, such as those from the Erlenmeyer collection, reflect their promotion.

The 1976 Cycladic exhibition at Karlsruhe itself contained a high percentage of objects without secure provenance or coming from illicit excavations such as Keros. The pattern of security of context for the Karlsruhe show is given in Table 5. Of the Cycladic figures, more than half are of unknown provenance, and scarcely more than one-eighth of known provenance. For the show as a whole, well over half the items are of unknown provenance, and less than a quarter are of known provenance.

Since the Karlsruhe loans were largely from Europe, and the Virginia loans from north America, we find telling the strong pattern, evident in both exhibitions, that recently surfaced finds of insecure provenance have come to dominate the Cycladic corpus in two continents.

During the 1960s, some mitigation of the loss of information was provided by archaeological salvage of what looters had overlooked or left in their holes, on Herakleia, Keros, Melos, Naxos, including figurines and fragments.

Is it possible to quantify the damage caused by illegal excavations? We think not. No cause has been offered to contradict Broodbank's view on the plundering of Dhaskaleio-Kavos – 'most dismal of archaeological prospects: the virtual bomb-site incongruously overlooking the brilliant blue sea, with the wreckage of material spread over the churned-up slopes of what was once the most extraordinary location in the Cyclades'.

New manufacture

A natural context for forgery is the emergence of a demand which no conventional supply can meet. Interestingly, the Cycladic supply did indeed change along with the demand: it was 'the 1960s, especially, [which] saw a burgeoning in . . . the availability of objects of high quality, primarily on the European antiquities market'.

Table 6 summarizes the period of acquisition of Cycladic figures into some major museums; for private collections, whose holdings are very much larger, no systematic information can be collated.

It is remarkable that the best pieces were reluctant to appear before this late date. Could it be, as Getz-Preziosi has noted, that this phenomenon is linked to the higher incidence of forged statuettes? In 1971 H. Hoffmann offered the following advice to prospective proprietors of Cycladic figurines:

'A New York dealer confided a "foolproof" test for Cycladic idols to me, and I shall pass it along for what it is worth. Hold the object between thumb and forefinger and strike it lightly on a doorsill, like a tuning fork. A forgery will emit a clear bell-like ring, whereas a genuine idol emits a dull thump. It is, of course, necessary to experiment with both genuine and imitation idols in order to accustom one's ear to the proper sound. (I must admit that I have not yet mastered the technique).'

The Karlsruhe exhibition drew attention to the growing problem of forged Cycladic figures. As early as 1967, J. Thimme had grown suspicious of 'a group of 'Cycladic' idols whose authenticity he doubted on a number of grounds: the figures either were of exceptionally large size – up to 120 cm in height – or else represented rare forms (e.g. double idols and musicians) which commanded especially high prices. Their surface was of peculiar appearance, they were all broken at head and knee, and they had been repaired with the same adhesive. Various aspects of the figures were studied with various techniques – features of execution, identification of the stone, geochemical analysis, surface encrustation, x-ray diffraction analysis and infra-red spectroscopy, ultra-violet light. Among further figures examined after 1969 was a group of three 'unbroken and unusually refined figures which one was inclined to consider too beautiful to be genuine'.

Thimme and Riederer's methods of twenty years ago were intended to develop a knowledge of surface encrustations that would authenticate ancient marble objects. Others we have consulted who have also explored the detection of fake marble objects have less or little confidence that fakes can be made to announce themselves in this way; marble figures in the early 1990s remain beyond physical-science methods of authentication – famously the 'Getty kouros' that is suspected of being fake, but has not been decisively authenticated nor decisively unmasked. We are reliably told that the surface of modern Cycladic figures can be enhanced to a helpful 'age' by wrapping in wholewheat spaghetti, whose strands leave marks happily like those left by roots in the ground. Mild acid is also commended. A celebrated Cycladic collector, we are reliably informed, was persuaded that the appearance of a figure under infra-red radiation would show if it was genuine, and added an infra-red lamp to the collecting equipment on the yacht.

The known period of busy faking is the 1960s. Doumas notes that fakes were so numerous – or fear of fakes so prevalent – as to reduce the level of looting of the Cycladic sites in the 1970s. Hood wrote in 1978, 'the problem of forgery cannot be ignored in considering them. Demand has clearly exceeded supply, and modern workshops in Athens and Naxos are reported as happily responding with figurines made of the same island marble'. The ready modern availability of marble from the various islands also ensures that typing of the marble sources by chemical signatures distinctive to an island or deposit, as has been attempted, would not assist in recognizing fakes. Getz-Preziosi was careful in her study of the presumed sculptors: 'I have resisted all temptation to reproduce figures of unusual types or unusually large size – forgers' favorites – that I have not had the opportunity to study "in the flesh"'.

Getz-Preziosi goes on, 'Even so, one can expect a skeptical reaction to some of the more unusual pieces, at least, on the part of some readers, particularly those who find it difficult to accept as incontestably genuine sculptures that do not come from documented excavations or whose pedigrees do not extend back into the last century.' The number with history known back to the last century is, we have seen, vanishingly few. Unhappily, we cannot presume that objects acquired before the 1960s boom, or even before the more modest collecting pre Second World War, are immune from suspicion. Take, for example, the small figurine in the Fitzwilliam Museum, acquired in 1934 and said to have been found on the island of Ios. The type is rare, and its genuineness has sometimes been brought into question. The body is similar to a large figurine now in a North American private collection (and on loan to the Cincinnati Art Museum) which is also said to have been found on Ios. That figurine had surfaced by 1925 when it was displayed at the Metropolitan Museum of Art in New York. It is especially close to the Fitzwilliam piece in the detail of having the hands raised diagonally on the chest. A third figure with breasts like the Fitzwilliam figure and also having the hands brought up to the chest is now in a private collection; it has no known provenance. This last falls into a group of six figures, the so-called 'Goulandris Hunter/Warrior Group', of which the other five are in the Goulandris Collection in Athens. Four are said to have been found in the same grave, but Getz-Preziosi has thought this unlikely as at least one 'was probably carved several generations before his time'. Even the find-spot has been disputed: Naxos, Spedos on Naxos, Phionta and Keros have been mentioned. Thus the parallels do not provide secure archaeological evidence for confirming authenticity for the Fitzwilliam figure. This inescapable doubt over an item which has indeed been in an 'old European collection' for nearly 60 years shows how an attractive route to authenticity, to have faith in the items known since before the market in Cycladic prospered so, in fact offers no security.

Fear arises especially for those classes of figures of which not one example comes from a secure context. For these, inescapably, there is a substantial chance that the whole class is bogus.

Most Cycladic figures, if gendered, appear female. Male figures (Table 7) are less common, Getz-Preziosi calls the male warriors 'a firmly established, if rare, recurring type'. Rare, yes – but how firmly established? The recognized male classes are standing male figures, warriors, and musicians.

All 13 standing male figures, of five types, are of unknown find-spot or, at best, 'said to be'.

The warrior figures are 'said to be'; fortunately, the old record which provides an early record may encourage one to believe the class is genuine.

The 'musical figures' of harpists have also been a matter of controversy. Eight are known in good preservation.. The pair in Karlsruhe are 'reliably reported' to have come from Thera; they are recorded as early as 1838. The example, now in the National Museum at Athens, said to come from a tomb on Keros where it was found with a flute-player was first reported in 1884. One in the Metropolitan Museum of Art in New York was acquired in 1947. Renfrew has had 'grave doubts' about its authenticity, although others have felt that it is genuine. and the pair in the Shelby and Leon Levy Collection are said to come from Amorgos. One now in a North American private collection surfaced on the London antiquities market in 1971, and the J. Paul Getty Museum acquired one in 1985 from a 'private collection' which had been said to have been found on Amorgos. A fragmentary example has been found in the Aphendika cemetery on Naxos; Getz-Preziosi notes that this is the only Cycladic harpist 'found in a controlled excavation'. Despite the sketchy archaeological contexts, Getz-Preziosi declares the authenticity of these eight to be 'assured', and proposes that these harpists were 'produced over a period of perhaps two hundred and fifty years by unusually confident sculptors'. This unusually confident estimate may create its own puzzle: how the type came to be known to the later sculptors when the earlier examples had gone away into graves.

A particular fear surrounds the larger, grander figures (Table 8), so distinctive, sought-after and expensive, where the cash stakes are so much higher. Not one of those over 50 cm comes from a secure archaeological context, and it is disconcerting that this class of figures preferred to wait until the market boomed before choosing to surface. Someone, somewhere is laughing all the way to the bank. The question remains, is it a looter or a faker or both? Consider the best course of action for a forger, or his representative, who wishes to be well rewarded for his work. A routine piece, a folded-arm figure of conventional small piece, will not stand out, but it will not command a special price. A figure of unique form or uniquely large size risks discovery through the interest it will raise; its special price may not cover the special risk. The best course may be to push 'at the edge of the envelope', to make a figure which is special enough to push up the market-price, but not so special as to be scrutinized with suspicious care. And once one and two and three and four extra-large figures, or extra-delightful musicians, have been taken into the corpus of accepted figures, the way is open for more of these special forgeries to be welcomed by precedent.

If the large figures are all genuine, the chance of a large figure being recovered from an archaeological context remains. Until then, the large figures are a puzzle. The small figures, found in graves and at settlement sites, are regarded as grave goods that previously had a function during life, perhaps in some household cult, suffering breakage and repair. The big statues, up to 140 cm or 148 cm tall are suggested by Renfrew to have been used in some more public cult, perhaps in a separate shrine. Without evidence of context, the idea is obliged to be a 'pure hypothesis'.

Renfrew has observed that ears, otherwise not provided on Cycladic heads, exist on some of these unusual large figures. One's confidence in the observation as a truth of prehistory is – again – necessarily at the same level as one's confidence in the authenticity of the figures.

The history of faking, as suspected for Cycladica, disconcertingly resembles the known story of the Neolithic Hacilar figures from southwest Turkey, which followed the course of discovery, of publication that brought its delights to collecting attention, a looting of the cemetery where fine items were to be found, and then of a mass production of fakes, many at Hacilar itself, using local materials and convincingly 'aged'. Many museums were deceived, including the British Museum and the Ashmolean before TL tests resolved the Hacilar collections into ancient and modern; 48 out of 66 Hacilar pots tested at the Oxford Research Laboratory were modern. The Cycladic story has been obliged to take a differentturn, as no comparable technique can recognize faked marble carvings.

– Table 9 here –

Drifting provenances

Comparison of information about pieces exhibited more than once shows that Cycladic provenances are subject to 'drifting', as stated find-spots change with time. Documentation, new information or reason for the drift is not given. These transformations are illustrated by the Karlsruhe (1976) and Richmond (1987) exhibitions (Table 9). The Karlsruhe exhibition displayed some twenty or so fragments which were reported to come from the looted deposit on Keros. The Richmond exhibition catalogue, which highlights the looting of Cycladic sites in its foreword, chooses to play down this provenance. Sixteen pieces in Richmond are 'reputedly from Keros', an unforthcoming remark as to just where they came from on Keros, and just when they were found. Yet the bibliography of one, no. 73, plainly shows that it was in this looted deposit, and two (Richmond nos. 75 and 82; Karlsruhe nos. 178 and 173) had been published in the Karlsruhe catalogue as 'From Keros - part of a large deposit', the orthodox euphemism for the Keros hoard. Indeed, another fragment of the hoard had trickled out between the exhibitions; Richmond no. 75 (originally in the Erlenmeyer collection), was headless in Karlsruhe, and had its head back when it got to Richmond. A head in the University of Missouri-Columbia Museum of Art and Archaeology was 'Provenance unknown' as Karlsruhe no. 177; in Virginia it was no. 81, 'Reputedly from Keros'. A provenance also surfaced for an abstract figure in a private collection: Karlsruhe no. 41 was of no provenance (ownership stated as Houston, D. and J. de Menil Collection CA6253); in Richmond it was no. 8, 'Reputedly found on Naxos' (and Private collection, acquired in 1962). A marble spool pyxis in the same collection also gained a Naxos provenance between Karlsruhe and Virginia: Karlsruhe no. 343 (Houston, D. and J. de Menil Collection CA5207) becomes North American no. 134 (Private collection, acquired in 1952).

Drifting find-spot also affects a remarkable group of marble figures, regarded as Neolithic in date, which first appeared in Karlsruhe. When the pieces – a sitting female figure, two squatting female figures with children on their backs, an animal and a bowl – were loaned to the Karlsruhe exhibition, nos. 4, 24, 25, 429, they were unpublished and in the possession of 'Switzerland Private Collection I'; they were said to be 'From Attica - part of a grave group', the knowledge that they were found together coming from 'a credible source'.). The group itself received a full discussion, and were 'subjected to x-ray analysis' which showed that 'the encrustation of all five pieces was pronounced to be typical of that found on genuine marble objects'. No specific find-spot for the group was mentioned more exact than Attica. When one of the figures appeared in Richmond, no. 2, by then in the collection of Shelby and Leon Levy, it was stated that the group were 'Reputedly found on an islet near Porto Raphti, Attica'. In 1990 the group was exhibited again, at the Metropolitan, and the history had changed again. The group was 'said to be from Euboea, or the east coast of Attica opposite, near Porto Raphti, perhaps from an islet attached to the mainland in prehistoric times'.

Normally, the find-spots of prehistoric artefacts (as distinct from their supposed places of manufacture) only move when clear information becomes properly available to revise the known history, or when decisive analytical evidence contradicts the supposed story. Not so with Cycladic figures. Of the Karlsruhe group, Getz-Preziosi states: 'While a Euboean or an Attic provenance is quite possible for this marble group, these are not the only places with rusty – that is, iron-rich – soil.' (p. 13) Indeed these are not. Most of Morocco has rusty, iron-rich soil. Without real information Cycladic students must listen to whatever 'said to be' is said to them, while the reasons that govern the 'said to be' are unknown; their scholarship is thereby reduced to clutching at rusty straws. And the politics of display have a clear rôle. If it is tactless to remind curators, owners and the public that items originate in the Keros hoard, a deposit that is commonly known to have been stolen and smuggled, then let the provenance be reticent; 'reputedly from Keros' will do nicely. One may see a pattern in attributions over time that amounts to a kind of fashion; in Bent's day figures were 'said to be' from Antiparos, now they are 'said to be' from Keros'.

Re-emergence of forgotten collections: a convenient fiction

When travellers like Theodore Bent brought back Cycladic figurines, interest in them was limited – even if that limited interest did extend to their notice by discerning collectors. Two in the Ashmolean once belonged to the Rev. Greville J. Chester, an important 19th-century academic collector. Many of the donors for British museums acquired the figures with the intention of conveying them into a museum. Notice which were the museums where Cycladic figures went early on – the British Museum, the Ashmolean and so on: they were never curios in the provinces. In the new century they had an artistic importance to the great sculptors. Yet it was only after the Second World War period that Cycladic figures came into artistic fashion as 'collectables' of importance. Nevertheless, one is given to understand that figures in North American collections came from European collections of the previous generation; as Getz-Preziosi plainly says of the general pattern, in the remarks quoted above, 'Many of them have come to the United States from European collections.' The Erlenmeyers' indeed was a European collection, formed 'between the late 1940s and the early 1960s' but Getz-Preziosi indicates that a very large part of it came from the illicit excavations on Keros; and if part comes from illicit excavations then which part of their collection is safe to acquire? Greek laws to protect antiquities go back well into the 19th century; a collection begun in the 1940s may be 'old' in the perception of a contemporary collector, but it is not old enough to pre-date the Greek legal process.

Perhaps the odd Cycladic figure, picked on a grand tour which went beyond the more obvious sites of classical Greece, did find its way to some forgotten attic, where it can one day be 'discovered' and re-emerge on the art market. At least three Cycladic figures are in fact documented to have 'sunk' from view during the 19th century, two folded-armed figurines from Naxos, whose whereabouts was unknown as early as 1888, and the Caria harpist. But it is hard to believe that many hundred Cycladic figures took this route – and all in such an obscurity that none of them were celebrated, or even recorded, in the catalogues and the sale-catalogues of the 'old European collections' before they moved into newer hands.

The place of the 'old European collection' as a source for Cycladic figures follows that of the wider antiquities market. Consider the patterns. Everyone involved with, and knowledgeable of, the business is aware that a great many objects are looted and smuggled; they are to some degree 'hot'. A small number of these are stolen from museums or other places where they are published or publicly known; in that way, the Kanakharia Byzantine mosaics from Cyprus and the Roman Medusa mosaic from Sparta museum were in recent years identified, and safely re-claimed by their proper owners. But these are unusual; most antiquities that are for sale have never been published, accessioned into a museum or come to public scholarly attention. Who can report a Cycladic figure as stolen when it has been lying unseen in a grave for more than 4000 years? And which agency is to investigate its movements between the ground and the celebration in a Swiss or a North American private collection or the London sale-room, when neither Swiss nor American nor British legal practice takes note of the unlawful acts under Greek law which its recovery and export involves? Some figures are fake. Because marble figures are not amenable to an analytical determination of age, there are no statistics for how many marble items have been tested or found wanting, as there might be for ceramics. The Oxford University Research Laboratory for Art and Archaeology, which authenticates ceramic objects by thermoluminescence, has reported 40% of the ceramic works submitted to it are not ancient. And what proportion are smuggled? Geraldine Norman has many years' experience in reporting the London art market for The Times and the Independent: she says, 'Eighty per cent of all antiquities that come on the market have been illegally excavated and smuggled.'

In November 1990, the science journal Nature published a colour illustration of a feather tabard of likely date between 200 BC and AD 800, on sale at a London dealer. It comes from the Nazca culture of desert Peru, an area notorious for looting and smuggling. On one of us enquiring, the dealer said the items were indeed of good provenance; they came from a 'collection'. How often does one here those phrases! – 'from a collection', 'from an old collection', 'from an old European collection'. But just which collection? Where are all those old collections? Why were the treasures they contained never known to the other old connoisseurs? Ricardo J. Elia of Boston University, recently remarked, 'the dealer's usual explanation about coming from an "old European family collection" has become a trite cynicism since it was exposed more than 15 years ago by Karl Meyer.' The phrase has now so lost meaning that it has now begun to fade from sales of many classical antiquities. The most celebrated antiquities sale of the year 1990, dispersing the Hunt brothers' vases and bronzes, included 53 lots; just one of these, in the Brummer collection (Zurich) by 1922, was documented as from an old European collection; of the rest, two had surfaced in the 1960s, and the other 50 yet more recently. There was no mention of any 'old European collections'; the source of the other 52 was unexplained. Will the idea persist for Cycladic antiquities, even less credibly since connoisseurs' esteem for them is so recent?

Wish-fulfilment

As well as complete figures, the Cycladic corpus includes a great many fragments that are recognizably portions of incomplete or broken figures. Even some complete figures are of simple form; the 'violin' variant of Cycladic figure has no distinct head, elongated neck, broad shoulders, narrow waist, broad hips and a rounded legless base, so it can only be recognized with reliability as depicting the human body because it fits the larger Cycladic pattern of abstraction from the human form. But simplicity and abstraction, at an extreme, makes a sculpted figure indistinguishable from a stone that was worn or shaped by human hand for reasons other than representation, or even from a pebble worn and shaped by natural causes alone. Without a documented archaeological find-spot and context, it is not always clear that the form alone reliably identifies the object as a figure fragment or other Cycladic antiquity.

Now taken into the Cycladic corpus, therefore, are shaped pieces of stone that may not be sculptures, but objects that modern possessors, whether purchasers or vendors, wish to treat as sculptures. That is what they then become.

Material consequences

Impact on the archaeological context of Cycladic figures

The small number of figures which have been recovered from secure archaeological contexts in itself raises questions about the damage inflicted upon the archaeological record by illicit excavation. If we take Getz-Preziosi's upper figure of 1600 Cycladic figurines of all types, then we can gain some idea of the number of graves which have probably been looted to service the demand by museums and collectors. Since figures furnish only a proportion of the graves, each figure may represent a certain number of looted graves, and some estimate of that number can be made from those archaeologically excavated cemeteries which record the ratio between numbers of graves and numbers of figures(Table 10).

– Table 10 here –

Not all of these 1600 figures came from graves. One can attempt some arithmetic for the consequences. If we remove those: from the Keros deposit, perhaps some 350; from excavations on Keros, approximately 50; from Ayia Irini on Keos, some 43; and from Phylakopi on Melos, recorded as 11 by the excavators; then some 450 are not from graves. This leaves 1150 of the corpus known in 1987 as probably taken from graves. Of these, only some 143 seem to have come from the controlled excavation of more than 1400 graves. This works out as approximately one figure for every ten graves. Using this ratio, the 1100 or 1200 figures without secure provenance may represent the destruction of some 11,000 or 12,000 graves, the vast majority without any figurines. Thus some 85% of this funerary record of the Early Bronze Age Cyclades may have been lost through this unscientific search for figurines. And, of course, for that large portion of the graves with figures which were emptied by looters, we have only the figures, or other saleable artefacts, without knowledge of their association or context.

One of the best-explored islands of the Cyclades is Naxos. Some 577 Early Cycladic tombs from ten cemeteries have yielded some 75 figures, a ratio of about one figure per eight graves. However not all cemeteries on Naxos yielded figures, in part because they had been already looted. Thus for every eight or nine Early Cycladic graves opened, a single marble figure has been redeemed. A very similar figure may be obtained for the island of Amorgos. The island of Antiparos was the location for some of the earliest excavations by Bent; if we take this island with the neighbouring ones of Paros and Despotiko, some 280 graves have yielded approximately 58 figures, one figure per 4.8, or nearly per five graves. A greater imbalance between number of graves and the numbers of figures was found at Chalandriani on Syros where only six figures were recovered from more than 600 graves, one figure per hundred graves. These figures suggest that for the islands of Naxos, Amorgos, Paros, Antiparos and Despotiko, each figure represents approximately seven graves opened; on adding Syros to the numbers, each Cycladic figure with a secure provenance is found to represent some ten excavated graves.

The estimate of the number of graves looted as some 12,000 has a large uncertainty; even a lower, and less insecure estimate of 10,000 graves makes a figure with large implications for Cycladic archaeology. This number would be the equivalent of 16 cemeteries of the size of the one excavated at Chalandriani. However this cemetery was exceptional in its size; it is one of only three which contained 100 or more graves. Even if one large cemetery of say 100 graves has been rifled, the remaining tombs may represent 500 to 660 small cemeteries of 15 to 20 graves each – a more usual figure. Destruction on this scale would be sure to distort the picture of settlement in the islands. This arithmetic shows that one cannot accept Pat Getz-Preziosi's comments in the Epilogue of her Sculptures of the Cyclades: 'Surely the Cyclades have many secrets still to be revealed.' Rather, every sign points to the obvious cemeteries in the Cyclades – which are small islands with much barren rock and shallow soil cover – now having been in large measure quarried out. Most of these figures are of the Fold Arm Type and thus belong to the ECII horizon (some 300 years), a phase with one of the more intense levels of occupation. Melos, for example, is thought to have had a population of only some 400 at this time, and estimates for the whole Cyclades in Early Cycladic II times include 6600 and 34,400.

These figures presume an even distribution of figures throughout the Cyclades. However it seems that in the Early Cycladic I horizon, most figures have been found on Naxos and Paros, and in the Early Cycladic II horizon on Naxos and Keros. Some islands which lacked marble sources, such as Thera and Melos, rarely seem to yield figures in funerary contexts. If this is the case, then the 10,000 graves must be located on even fewer islands. Getz-Preziosi notes, 'the distribution of marble objects within the Cyclades has been badly blurred by extensive illicit digging and the corresponding paucity of documented finds'. Some pattern may survive in the blurring: the illegal digging is in pursuit of antiquities, and must generally follow the looters' experienced knowledge of where the goods are to be found. Another effect can arise from accidents of geography, and the modern trading ports. At one time Keros belonged to the Panagia Hozoviotissa monastery on Amorgos; in addition, Keros is far nearer to the main town on Amorgos than it is to the capital of Naxos, way up the west coast. Amorgos forms a natural outlet during the early stages of discovery and dispersion.

These known, or perceived, patterns do not always fit such information as is offered for individual figures, and the literature therefore sometimes provides an alternative to a 'said to be', when the 'said to be' is thought unconvincing. A headless and legless marble figure formerly in the Erlenmeyer Collection, was once said to have been found on Amorgos but the sale catalogue comments that 'it is quite possible that it was actually recovered on Keros'.

Context is important for the most schematic figures. Take, for example, the three pieces of marble sold at the Erlenmeyer sale, December 1990, as its lots 134–6. Published by the Erlenmeyers in 1965, they had been 'authenticated' by appearing in the Karlsruhe exhibition of 1976, where they were described as 'abstract-schematic idols'. One is a conical piece of marble 9 cm high, identified in the 1990 auction catalogue as a pestle, although the description is more cautious. The second, a sub-rectangular block, also 9 cm, is identified as a baitylos [sacred stone]; the description is again more cautious. The third, a sub-trapezoid pebble 8 cm high, which to our eyes has very little in common with Cycladic representations of the human body, is called a 'highly schematized marble figure'; the description, again more cautious, only claims that the 'small object of simple form and strong tactile appeal resembles the torso of a normal Cycladic folded-arm figure'. The object does to a certain degree resemble that shape, but so does many a pebble that lies along any of the shingle beaches of the world radiating the strongest tactile appeal to the eye of the passing connoisseur. It is well known that marble pebbles were placed in Early Cycladic graves, but without a context they are no more than pebbles. They can only be identified as artefacts of a human interest by a secure knowledge of that context. When that context is lost, they lose their standing and are, once more, beach pebbles. The Erlenmeyer 'abstract-schematic idols' are not known for certain to come from a grave, because their history is unknown. Without history, they are become beach pebbles once more. The same is true of the 'knob (?)' in the George Ortiz Collection in Geneva, an object of no known provenance whose declared history makes no proof that it is a genuine Cycladic antiquity. Yet it is seen as art: 'The form of the object is perfect, the workmanship of superior quality. It is at present unique. What its purpose was is not known.' These objects by their lack of history are reduced to meaningless curiosities, 'things sold under the belief they are works of art'.

A possible corollary to these revered objects without history is found in those objects unrevered because they lack sufficient distinctive form. What is a looter to do with beach pebbles or 'abstract-schematic' figures found in a grave that do not by their form sufficiently announce themselves as Cycladic antiquities? Asthe form is not diagnostic, and as the history – the only true proof for items of these undiagnostic forms – is untellable, we fear objects of these kinds are left in the grave or otherwise discarded by those without the authority to declare them to be of a perfect form.

Finally, one may notice the effect of a changing climate for the open sale of classes of antiquities, like Cycladic, where there is some fear that items may be of unreliable history, whether through faking or through illicit export. If no account of the history of some pieces can be given that would be satisfactory with respect to those fears, then a discreet way forward may be to give no explanation at all. If silence is in order for some items, then reticence may be a prudent habit in general. Certainly, one can notice that published sales catalogues of recent years seem to be less forthcoming about the history of items than was once customary.

– Table 11 here –

Table 11 summarizes the accounts offered for the 40 items in the selling exhibition at Galerie Heidi Vollmoeller, Zurich, 1989. No account is offered for 34. A history of the other six is offered as far back as the Karlsruhe exhibition and catalogue of 1976; two of these were already in the gallery's ownership at that time; the history of none goes beyond Karlsruhe.

Consequences for chronology

Insecure provenances are common outside Cycladic. Some essentials of classical chronology are, like Cycladic, dependent on 'said to be' and the vagaries of mobile collections. The key pieces of pottery in the study of early Greek pottery from the Black Sea are a MGII hydriske purchased from a dealer in 1909 and 'said to be' from Berezan (yet no examples of such pottery have been found from excavated contexts), and LG kotyle and kantharos fragments in Museum of Classical Archaeology, Cambridge (perhaps from Istria but possibly slipped into the wrong drawer from the Al Mina finds).

As the context for so many Cycladic figures is lost, the framework depends on the 10% of the corpus which has been found in secure archaeological contexts. A figure without context can only be dated by reference to other figures which have a secure provenance. With the present evidence a schema has been created to trace a progression from the Abstract through the Hybrid and the Precanonical to the Canonical and then on to the Postcanonical of the Chalandriani type. However attractive this theoretical outline might be as an abstract of an artistic cycle, it is necessarily remote from the evidence that ought to be available. Are Precanonical figures consistently found in early contexts? Are Postcanonical consistently in late? Are different types contemporary, or do they follow in a distinct succession? If they overlap, do the contemporary types fit with the scheme of things? The other 90% of the corpus might have answered these questions but their evidence has been lost for ever. Barber comments, 'our understanding both of the chronology of types and of the ways in which these objects were used, is slight by comparison with the total number of finds, though just enough figurines have been found in systematic excavations to provide a key'. We think Barber's view of the weight of evidence is too generous; we are not convinced that 'just enough figurines have been found in systematic excavations' to carry the schema which has been built upon it. Within the Keros-Syros phase, Broodbank notices, the case for an increasing 'abstraction' depends on a relative chronology for the figure types which derives entirely from a priori evolutionary assumptions, and supported by neither stratigraphy nor seriation; this proposition seems to relate to no empirical information at all.

We fear that ideas about Cycladic chronology have now come to depend too much on preconceptions about the range of types and variants that ought to be in production at any one time, and about the stages through which those types and variants ought to progress in a highly disciplined aesthetic order. The figures which do not fit are set aside as late, decadent or aberrant, so as not to challenge the studied truth of the aesthetic. Those elegant schemas sort and order the figures into a chronological pattern, a pattern that can then be depended on as confirming the historical truth of the schemas.

'Workshops' and regional variation

A further problem is the attempt to isolate the areas where the different 'sculptors' were operating by comparing find-spots. Of the eight 'classical sculptors' and the five 'late classical sculptors', Getz-Preziosi produced a pattern of distribution of the works. Seven different islands are represented: Paros, Naxos, Keros, Amorgos, Keos, Ios and Crete. Of these, Paros, Amorgos and Ios have no secure findspots, and Naxos and Keros dominate those with secure ones. This lack of secure provenances and archaeological contexts deprives scholarship from the opportunity to discover if the Masters are masters or whether in fact the groups stand for regional styles, a point made by von Bothmer. There is an equivalent question by way of distribution over time: if there was good evidence of context and therefore of date, one could hope to establish if the works of a master indeed were contained within a single human life-span.

– Table 12 here –

Size and function in the figures

One question about the use of the figures has focussed on whether or not the large examples were used in effect as cult-statues whereas the smaller ones were usually confined to graves. The range of size for the works attributed to Getz-Preziosi's sixteen 'sculptors' is a guide (Table 12).

More than half of the ECII figures are less than 40 cm high, and more than 80% less than 60 cm. Of the tall figures, those supposed to be cult-statues, not one has a secure provenance, although one is said to come from Amorgos.. Getz-Preziosi comments, in respect of size, 'the purchaser of a piece probably neither commissioned it to conform to a precise measurement nor would he have paid for it 'by the centimeter' [sic], a method of pricing rumoured to have been in use among grave robbers in recent years'. Again, it proves hard to find certain knowledge in a mass of uncertain reports.

Intellectual consequences

Reduction and corruption of the factual record: towards an impossibility of understanding Cycladic prehistory?

A first and immediate consequence of the circumstances under which some 90% of known Cycladic figures have been removed from the ground and moved round the world is a total loss of reliable information as to their archaeological context. For those whose archaeological context was a tomb like those known archaeologically, the loss is quantitative rather than qualitative. There remain a variety of possible non-funerary contexts for figures, that have not been available to archaeological study: which classes of figure, if any, come from sanctuaries or other types of sacred places? And what kind of deposit is the Keros hoard – an immense tomb- or cemetery-group? a sanctuary deposit? or something else again? Conversely, there are some types of figure unknown or barely known from archaeological contexts, such as the harpists; we therefore do not know if that class of figure comes from tombs just like the others, or whether there is a different pattern of context, to go with the different form, and reflecting a different prehistoric purpose and meaning. The corpus of figures is vastly enlarged by the recent surfacings, but the information as to context has scarcely changed. And for every tomb that was destroyed and yielded a figure, there were perhaps 10 destroyed without even producing that.

Many figures have surfaced with some declared provenance, to the island of Naxos, say, but without supporting evidence – understandably if to document its history is to admit to improper acts. These corrupt the record, for it is in the interests of the seller to give information that the buyer will be happy to hear, and which offers no clue by which the buyer – or any agent of the Antiquities Service – can trace back the figure's movements from the earth. Forgeries corrupt the record whenever they are declared as having a provenance. The Keros hoard offers particular opportunity for the forger, as it is a shadowy collection, understood to be very large and diverse, never fully or securely documented. The hoard provides a convenient alibi under which a forgery, whether orthodox or novel in its form, can be offered as both plausible and impossible of being authenticated. Given the chain of intermediaries through which a figure may pass between the hole in the Cycladic ground and the connoisseur's cabinet – none of whom have necessary good cause faithfully to pass on what they hear – one can have no confidence that any knowledge of provenance that the figure may chance to carry at the start of the chain will in fact arrive unchanged at the end. The phrase 'said to be from . . .' that one sees in the catalogues, means just that. One must ask, as a historian does of the words a document offers for study, Said to whom? by whom? when? with what purpose?

Nevertheless, the connoisseur, concerned to be certain that his purchase is what it is supposed to be, has good cause to wish to know that provenance to be correct, for the same reason as he wishes the object to be genuine. It is also in the interest of the dealer of integrity to offer reliable information. Unfortunately the circumstances of sale, whether in an open market or privately, rarely offer any full reliability. The recent example of the Sevso treasure – a hoard of late Roman metalwork put on sale as coming from Lebanon and with a Lebanese export certificate, but claimed also by the Yugoslavian government to be from its territory, and 'reliably rumoured' ('said to be') also to come from Hungary – is but one example that shows why declared provenance is a matter subject to many influences beyond the real place where an object first surfaced.

The same corruptions of reliable knowledge, arising from the existence of uncertain and false objects with uncertain and false contexts and provenances, afflict all classes of antiquities. Take an example that could be repeated from any one of many recent sale catalogues, 'Two Danish flint axes, Neolithic period, circa 2500–200 BC, one from Store Jyndevad . . . Another, circa 2500–2000 B.C., from Alslev v. Hjordkaer.' Perhaps, even probably, these axes are what they say they are: a reliable history may be available from the previous owner; some Danish axes are thought distinctive in their shape and in the appearance of the flint, providing confirmation of the declared provenance; there is little risk that an agency fearing illegal or unethical elements to the history will take a suspicious interest. No obloquy will go with their sale, no visits will be made from the Danish Embassy to the auction-house, no injunctions will be sought from the High Court. A great many polished axes from the European Neolithic were privately collected in the last century, so there are indeed many 'old European collections' in which the things can be found in quantity, and polished axes are common stray finds across northern Europe to this day. Yet our corpus of finds with documented and reliable contexts and provenances for the Danish Neolithic may seem sufficiently large, and our knowledge of the Danish Neolithic may appear sufficiently broad-based, that it might be able to withstand the loss of knowledge that could once have gone with the quite large number of flint axes and other artefacts now in the older, ill-documented museum collections and in circulation among the connoisseurs. The same may appear true of some classes of classical antiquity, despite so many centuries of collecting before archaeological excavation in the modern sense began: those that surface now are 'just more of the same'. It is certainly not true for the figures from the prehistoric Cyclades or for our understanding of Cycladic prehistory.

– Tables 13 & 14 here –

Although esteem for Cycladic antiquities is a recent phenomenon, and with it looting on the larger scale, it has the appearance of affecting a greater proportion of the total corpus than is usual for classical antiquities. In the absence of any published statistics to show what proportion of the classical antiquities in major museums and private collections are of a secure history, the point can only be illustrated on a small scale. The 1979/80 loan exhibition, Greek art of the Aegean islands, the first ever sent by the Republic of Greece to the United States, brought together 193 items from the Cyclades, early and later, with material of Mycenaean, Geometric, Archaic and Classic periods from Crete, Rhodes and other islands. The major collections lending were the National Museum of Athens, the Louvre and the Metropolitan. Table 13 shows the provenance listed for the items divided first by its ten sections, and then by the lending museums. Table 14 shows the provenance in the corpora for some classes of Archaic sculpture.

– Table 15 here –

The pattern that is visible largely follows from the date of acquisition. The supply of legitimate antiquities has dried up; as demand has continued, so it has to be supplied increasingly by illegitimate means. Private collectors of Greek antiquities in North America, as the newest class to pursue that demand, have naturally had to depend more on illegitimate sources. The proportion of unprovenanced or doubtfully provenanced items in some North American private collections of Greek painted vases is correspondingly as high as for the North American collections of Cycladic figures. Of the items from the Shelby White and Leon Levy collection that were exhibited at the Metropolitan Museum (New York) 1990 with much celebration and no reserve, an astonishing 94% are of unknown provenance (Table 15). When it comes to the European collections, which may be 'older' in the generality of Greek antiquities, the recent promotion of Cycladic figures into the desired canon of collectables shows itself again in the high proportion of unprovenanced or doubtfully provenanced figures.

B.F. Cook, Keeper of Greek and Roman Antiquities at the British Museum, recently set out his view of the market in classical antiquities:

'Given the known pillage of archaeological sites and the flood of unprovenanced antiquities on the market, there is clearly reasonable cause to believe that a considerable number, if not the majority, of these unprovenanced objects have been clandestinely excavated and illegally smuggled out of their countries of origin. Under these circumstances all unprovenanced antiquities are tainted with suspicion (cf. Chippindale 1990: 704), and objects considered for acquisition should therefore come with some evidence that they have not recently come into the market through the hands of criminals abroad.'

The British Museum policy, therefore is:

'Objects of foreign origin are only acquired if the laws of the country of origin allow it, a point cleared as a matter of routine in any report submitted to the Trustees on a purchase or gift.'

For this reason, Mr Cook explains, 'relatively few objects have been acquired by the British Museum's Department of Greek and Roman Antiquities in recent years'. The remarks apply a fortiori to Cycladic figures, and here is the problem for any collection, private or public, less favoured by chance and history in its assets than the British Museum. There is no supply, practically, of first-rate Cycladic figures that 'come with some evidence that they have not recently come into the market through the hands of criminals abroad'. If a museum or a collector decides a Cycladic figure must be acquired, then choice has usually to be made amongst unprovenanced antiquities that are variously tainted with suspicion. The unacceptable alternative is no Cycladic figure at all.

Art, eternal value and the idea of redemption

It seems to us a remarkable fact that the history of esteem for Cycladic figures, as we have documented it, has led to so many actions on, or beyond, the margins of legality. Esteem for so many other classes of antiquities has had such similar consequences that all this may no longer surprise or shock. As elsewhere in this paper, we make no moralizing comment and confine ourselves to observation. Yet consider the respected place in western society that the connoisseurs and the museums hold; these are indeed the pillars of society. The discrepancy is so striking between that standing – and the correctness of moral behaviour that goes with it – and the reality that lies behind 'no history' that one must look for motives by way of explanation that reach into the domain of the psychological.

What are those motives?

A first element is the wish to acquire. Sir Robert Sainsbury says of his collecting: 'I have, for over, forty years, been a "passionate acquirer" – a passionate acquirer of works of art that have appealed to me', and goes on to quote Thomas Hoving of the Metropolitan's more explicit statement: 'if something is truly grand it has a kinetic effect. Something akin to being physically impelled. It is, frankly, a charge, both cerebral and viscerotonic, not unlike a sexual experience.' The Metropolitan Museum press release for the catalogue of the White and Levy collection is headed with a quotation from its foreword: 'True and Passionate Collectors'. Going with this passion is the competitive framework within which collecting takes place. There are only so many Cycladic figures, or only so many great Cycladic figures, only eight – just eight harpists – and some of them are spoken for! Each that another has, I cannot have for myself. This is the old motive, 'the chase, the capture' that drives the spirit of acquisition by museum and by individual collector. Dietrich von Bothmer, Distinguished Research Curator of Greek and Roman Art at the Metropolitan Museum, remarks: 'The creation of the great museums of the world – the Metropolitan Museum of Art is no exception – and their continued existence and well-being owe much to a special breed of individual: the collector. No matter where or when they plied their trade, these individuals are linked by a common thread - the ardent desire to own works of art.' Alan Shestack, Director of the Museum of Fine Arts, Boston, says, in his important essay in The ethics of collecting cultural property:

'Museum professionals are acquirers; we are inherently greedy collectors. Most of us go into the profession because the desire to accumulate and bring together objects of quality is in our blood. We are personally and professionally devoted to adding to and improving our holdings That is what makes us tick. And to consciously or intentionally turn down a highly desirable object we can afford to buy on the basis that we suspect that it might have been removed illegally from its country of origin – and also knowing that it will end up in the collection of a rival institution or an unscrupulous private collector is a very hard thing to do.'

Despite these pressures, the museums are by degrees coming to follow a common policy that rules out the acquisition of newly-surfaced antiquities. In the museum world, acquisitions are now become 'subject to the constraints of committees, to financial restrictions, and to the relative academicism of many of their curatorial guardians'. In the private field, the spirit of chase and capture can still prevail. As the Metropolitan's Director puts it, 'in the assembling of private collections every whim can be exercised at will, and the preferences of the collectors tirelessly pursued and indulged: the result is fresh and often surprising because it reflects the collectors' personal taste and a sense of risk and adventure'. It is also pertinent that the curator, at the British Museum at least, is obliged to guarantee the good history of a planned acquisition, where the private collector has a wider freedom. Indeed, as museums become reluctant directly to be seen to acquire works of uncertain history, then the private collections – acquired outside public or trustees' gaze, unsupervised by an apparatus of formal codes of conduct, and then transferred to the museum – can instead be the means by which the great museums can satisfy their ardent desires. The pattern can now begin to be discerned for Cycladic of a route that goes from the ground through intermediaries and across frontiers to a private collection in a 'safe' country in Europe or north America, and then by loan, gift, bequest or purchase from private collection into a museum. Dividing the route into distinct stages in this way serves to distance the museum, as ultimate recipient, from those discrete transactions by which the stolen and smuggled object is laundered into a civic treasure.

In the case of Cycladic, there is a large possibility that the supply, by new manufacture or by surfacing or otherwise, could outstrip demand. A considerable time sometimes has been known to elapse between some objects becoming known and being available for sale; Cycladic figures and other antiquities, lent by Galerie Heidi Vollmoeller (Zurich) to Karlsruhe in 1976, were available at the gallery's stand at the Basel art fair in 1989. Some more of the 'Keros hoard' trickled out subsequent to the release of what had seemed its entirety. Cycladic market-makers are clearly aware of the danger, and adjust the supply accordingly.

Another aspect may be that antiquities, like other works of art, are one means by which mere money, even if vulgarly acquired, can become a fine proprietorship; the celebrated way in England for the rising new rich to meld into the old rich is to buy or to build a country pile, and thereby to join the aristocratic ranks of the landed gentry. A connoisseur's regard for fine old things makes the same statement. There is nothing necessarily special about Cycladica here; they are just another class of possessions to esteem. Yet some collectables are grander than others; enthusiasts for philately have to live with the phrase 'stamp-collecting' as a derogatory term for aimless collecting. We think we can perceive, in the rising tide of regard, a place for Cycladic not as just one more ancient class of art to go alongside Cypriot pottery, South Italian red-figure vases or little Etruscan bronzes, but as an art whose special and higher status is demonstrated by its remarkable aesthetic shared by the modern masters. It takes a special connoisseur properly to appreciate the refined quality of Cycladic, so cool and understated by the side of the more florid classical forms. For this reason, we expect Cycladic art to continue to rise in esteem and, by degrees, for its market value to overhaul longer-established classes of 'easier' classical collectables.

Although archaeology and art history are newly awash with self-conscious theory, not much is written on the philosophy of contemporary connoisseurship. J.T. Spike, 'an independent art historian specializing in Italian paintings', remarks while writing about a contemporary-art dealer's exhibition which included Cycladic figures, 'Questions of quality are not often addressed in writing as those responsible for much of what is written – art historians and archaeologists – regard works of art more as documents or facts than as objects of aesthetic appeal'. Arguing that we can enjoy things without their context, Spike says of the 'Cycladic idols':

'To return to idols, we need not apologize nor feel discouraged by the realization that we will never understand these ancient terracottas and stones with anything like the meaning that they held for the worshipper who reverently gave them as offerings. . . . Great art is large in spirit: it is sufficient and little short of miraculous, come to think of it, that these works of art can seize the imagination of persons inhabiting any part of the world in 1990 AD.'

All this is true and fair, in our opinion, provided it is clearly understood that the connoisseur's appreciation, in a place, time, and intellectual climate far removed from that of the object's maker, may intersect little or not at all with the meaning, function or vision that the same artefact held in its own, place, time and society. But does that justify the connoisseur of this eternal value in seizing the treasure without respect for the rights of persons inhabiting other parts of their world in AD 1992?

The conventional motives, like the two outlined above, do not seem so strong that they sufficiently explain the disregard of private collectors and of some museums for the actions which, by their purchases, they become accomplices to. Perhaps too little notice is taken of the power of a vision and a spirit of redemption in these matters. Consider. An ancient genius, gifted by providence with the skills to craft Cycladic figures of museum quality and graced with a society in which his genius was nourished and succoured, contrived enduring masterpieces out of the base rock of a tiny, dry, poor island. The figures, 'austere and stylized yet delicate and fully formed', which in their own time may have been idols, gods, demons or images of the dead, are eternal as Art. That is why, in the 1920s it was Moore and Modigliani, artists not scholars who rediscovered the incomparable beauty of these works, why Pablo Picasso viewed a Cycladic sculpture he owned as finer than a Brancusi. In ancient times these masterpieces came to be lost to the earth; they may in a deep sense have been lost even before they went back into the earth, if they chanced to fall into the possession of contemporaries unable to understand their Art, who might have even regarded them as functional little things of no moment, as mere idols.

Buried into inhuman soil, the figures may be lost forever. Yet for a fortunate few, there is the chance of redemption from that fall into earth, to be taken up once more out of the base clay of an Aegean island, to enjoy again an intimacy with a human mind and a human spirit that can reach out to its high and enduring art. What honour for a Cycladic figure could be greater than to be owned by Picasso or esteemed by Giacometti! This is the vocation of the modern connoisseur of Cycladic, to recover that which is lost, to esteem it for its greatness, and by that spiritual communion to make the figure come to glory again, and so to let the spirit of the master who made it live once more: master shall speak to master across so many remote centuries in the language of masterpieces, the shared language that needs no common tongue!

If the impulse behind the pursuit of Cycladic is of this nature, then several consequences follow.

First, the essence is in the fact of redemption, the bringing again from the darkness of earth into the light. The circumstances of redemption are a lesser matter. It does not matter much whether a figure is retrieved by archaeological excavation, by chance discovery or by archaiokapiloi; and there is nothing of special merit about an archaeological recovery over any other.

Second, it is then of the essence that the figure newly redeemed from the earth be conveyed into the possession of a person who shares the grace fully to grasp its aesthetic nature, to respond to its eternal virtue. A archaiokapilos does not qualify, although he will suffice as the vehicle by which the thing is redeemed. An archaeologist may suffice – but not if, instead of bringing the figure to light, he then consigns it instead into the modern darkness of some remote, closed store-room. Nor will he qualify if he overlooks the spirit of the figure and insists on a narrow, materialist attitude in which this figure of grace is just a physical object, some more unfelt data.

Third, it is immaterial where the redeemer chances to reside. These values are of the eternal, and may equally be found in Fort Worth (Texas) or Ludwigshafen-am-Rhein; they are as likely to inhabit Athens, Alabama, as Athens, Greece. After all, providence chose to endow the people on some dry, scruffy, barren Aegean islands with the ancient grace to create these spirits out of the pebbles on the beach. There is no special claim of descent or geography which privileges the claim of persons or authorities in the Cyclades, or elsewhere within the modern state of Greece, over those of other lands. Rather, the higher calling of the connoisseur, and the obligation to take the act of redemption to its completion, requires that the finest figures be taken before those rare eyes that can grasp their deeper meaning. It is in this way that the human spirit is communicated, from age to age, despite the spite and petty obstacles invented by those to whom it is not given to understand these deep matters. A different view of redemption is clearly held by the Greek authorities who seek the return of these figures from alien lands to the museums and collections of their native country.

We do not think the motivations sketched above fully account for the desire of some collectors and some museums to over-ride the legal and ethical obstacles to looting and export of Cycladic figures from Greece. But we do think that this is the spirit which stands behind the realities of the trade; we see it as explaining why little weight is given to the archaeological wish that Cycladic figures be recovered in a controlled manner that allows knowledge of context, and why the Greek government's efforts to keep Cycladic figures in Greece are regarded by some in the collecting countries of western Europe and north America as no more than a tiresome and parochial possessiveness.

It is also to be remembered that the figures were put into graves. Private collections, and even museum collections, of human bones are increasingly regarded as morally dubious. With good logic, some indigenous groups aspiring to recover and re-bury human remains they regard as their own have come to regard the grave as encompassing the grave-goods, and therefore require reburial to include artefacts along with the bones. On this view, the Cycladic figures should go back into the ground, except where they can be unambiguously traced to a non-funerary provenance, and may properly remain above the ground. One man's redemption is another man's grave-robbing.

Comparative moral values in these matters throw up some anomalies: Cycladica from the Erlenmeyer collection were auctioned to fund the Erlenmeyer Stiftung, 'a foundation for animal welfare', which funds such ideals as 'Save the Elephant' and 'Save the Rain Forest'. Another charity might equally choose to sell smuggled ivory and rhino horn in order to safeguard the welfare of Cycladic archaeological sites.

Shelby White and Leon Levy say of their collection, 'The excitement of collecting has been not only the joy of possessing a beautiful object but the fascination of discovering the links between that object and its place in history.' What are the histories of the objects in their collection, where are their places in history? Our Table 15 (page 00) shows one kind of an answer: 2 'known', 10 'said to be', 1 'perhaps' or 'possibly', 217 'unknown'; 230 total. With so much information lost in the discovery, export and acquisition of these beautiful objects, there are indeed many links between those objects and their place in history to be fascinatingly re-made as best can be done from the fragments.

Identifying the masters of Cycladic art: the available information

We have described, and attempted to quantify, the degree to which provenances of Cycladic figures are lost or may have been falsified. That lost knowledge damages the potential for recognizing pattern among the figures, whether in morphology, in context, or in distributions in space and time, and particularly afflicts Getz-Preziosi's identifications of individual masters among the Cycladic figure-makers. Getz-Preziosi's suggestion that her 'Goulandris Master' was a Naxian may be interesting but it is unprovable. Certainly a figure attributed to 'his' hand was found in tomb 23 of the Aplomata cemetery on Naxos, but others come, or are said to come, from Amorgos and Keros. Until such time as more of these figures are found in a secure archaeological context, and when they form a large percentage of the pieces attributed to the 'Master', ideas of this nature remain in the realms of unprovable speculation. We may have demonstrated, above, that destruction of graves is now sufficiently extensive that such time will never come.

The attribution of hands presumes that these figures were the works of single people. Yet there is a possibility that the picture emerging is one of regional styles and that instead of talking about, say, 'the Fitzwilliam Master', we should be talking about the regional characteristics of the island of Amorgos. However, so few of these figures have held their provenance (and alleged provenances are of little help) that these differences may now be beyond identification. Oustinoff's experiments in making Cycladic figures shows that the labour held in the known corpus is far from sufficient to make demands on an established corps of specialized full-time sculptors.

The corpus itself is not to be trusted, as we have shown. In consequence, an attempt to identify individual hands and even the characteristics of the sculptors within the corpus must be treated with caution. How are we to know if the works attributed to the Goulandris Master reflects 'l'expression condensée du calme, de l'équilibre et de l'harmonie des formes simples et complémentaires que le terme "idole cycladique" évoque aujourd'hui'?

Getz-Preziosi explains her purpose in identifying masters, 'A number of sculptors with easily recognizable styles are now known from two or more published works. In an attempt to rescue them from total anonymity and to facilitate identification, I have named the sculptors after museums (or museum locations) or private collections that contain one or more well-preserved examples of their work; occasionally I have used the name of an archaeologist who has unearthed important pieces attributable to a particular artist'.

The lack of secure provenance for such a high proportion of these figures must cast doubt on their authenticity. If different 'sculptors' (or perhaps regional styles) may be discerned, then confirmation of the authenticity of each 'master' (or at least part of the group) could be given by the inclusion of at least three figures with a confirmed find-spot. Of Getz-Preziosi's sixteen 'sculptors', six have the potential of not being ancient; however one piece attributed to 'The Metropolitan Museum Master' was said to have been found on Delos in 1860, and the 'name-piece' of 'The Ashmolean Master' is said to have been acquired on Amorgos at the end of the nineteenth century.

We address below the intellectual framework within which the identifying of masters takes place. For this portion of the paper, we do not dispute whether the masters did or may exist, or whether their work indeed holds an individual signature which the educated eye can grasp. Then, the question arises, 'how many works suffice to define a master?' A master who exists in a single work is not a master, but a single figure, as the point of the concept is to define an artist's style as evident in a range of work. Two might suffice, but make for an artistic personality too slight to be useful. We insist on three as the minimum number of works that make a master. Knowing the difficulty with fakes, we also request that those works be genuine, for what could be easier than to fake a more work in the style of a single genuine figure? No certainty exists as to what in the Cycladic corpus is good, but our best guide will be those figures that are of a secure documented context, for the most part an archaeological excavation. We therefore are obliged to set aside the works that are 'said to be', 'perhaps', 'possibly' or 'unknown' in provenance, as capable only of providing circumstantial evidence, and require that a master offer three or more works of a secure context.

– Table 16 here –

Table 16 sets out the works of the masters, as Getz-Preziosi recognizes them, together with their security of provenance. In it, we find two secure masters, the Doumas and the Naxos Museum; for each there are five works of secure context. There are eight masters we class as insecure because they offer only one or two works of secure context. And there are six masters we have, reluctantly, to call inadequate because they offer no works of a secure context; these are the masters whose entire corpus is built on provenances of sand. If we were to ask, not unreasonably, for 8 or 10 pieces as necessary to define masters, then even the two secure masters become insecure.

Identifying the individual craftsman in prehistory

By definition, no actual named individuals can be recovered from the material objects of a truly prehistoric context, and no prehistoric artists can be named. What can more often be isolated, perhaps, is a corpus of work with a range of distinctive and idiosyncratic traits which may appear to approximate to the range of variability of the work of a single individual, as has been done for Cycladic figures, and also for Benin bronzes and West African wood-carving. But it will never be clear that one of these 'analytical individuals' ('Master of the cow sacrifice', 'Master of the uneven eyes') actually is a single person, or whether the work represents in fact either less than a single craft-worker (whose work actually runs to a wider span of variation) or more than a single craft-worker, perhaps a workshop, a local or regional tradition, or a particular chronological horizon; or whether the distinctive idiosyncrasies actually arise from a particular function or purpose, or some other factor quite extraneous to the personality of the maker. Hill and Gunn's The individual in prehistory offers a range of studies in this pursuit for several prehistoric technologies. Its authors generally find the identification of an actual individual an unhappy ambition, since the best that can be managed will always be the analytical individual. In our opinion the success of the identification of unnamed masters in Renaissance painting does not indicate that the ambition, or the method, has equivalent value for prehistoric study. What is available for the Renaissance, and unavailable for Cycladic, is a documented body of work for many named artists, so one can judge the possible corpus of an unnamed master by reference to the range of the known Daddi, Gaddi or Botticelli; a prehistoric context offers no such prototypes for how an individual personality does or does not show itself in visible and distinctive traits of the body of material in question.

Redman's essay in The individual in prehistory is sceptical of the value, as well as the possibility, of advancing from the 'analytical individual' to the actual individual in prehistory. He remarks: 'I believe that a preoccupation with identifying individuals will lead to debates over the validity of identification displacing our primary concern – for the processes to be explicated with this information. All too often it seems as though scholars who study the behavior of particular individuals do so at the cost of more general investigations.'

We are less gloomy than Redman, who thinks his own scepticism 'may be excessive', because we can see one good route to testing whether an 'analytical individual' in prehistory may actually be one human being. Take a body of material, like Cycladic figures, which seem to show a variability of a kind amenable to identification of the individual. Identify within it an 'analytical individual'. Then examine the distribution in space and in time of the individual corpus, and see if that matches with the range to be expected from the output of a single craft-worker. In the Cyclades, as elsewhere, there are some obstacles. In a small archipelago known to have had networks of exchange and trade, spatial distribution may not be decisive. Chronological resolution is more promising, but may be blurred if figures are in circulation for any prolonged period, rather than going rapidly into the ground upon their making.

Resolution of this important intellectual issue is made impossible by the material consequences of esteem for the Cycladic Masters. As is shown above, the independent evidence for the distribution in space and in time of the figures is vanishingly small: 14 of the 16 Masters are insecure or inadequate (table 17), and the information available in respect of the two secure masters (the Doumas and the Naxos Museum) is not sufficient to address their distribution in space and time. The identification of individual Cycladic Masters, then, is a research proposal impossible of realization and therefore valueless to pursue. But the idea of the Cycladic Masters is central to the esteem now held for Cycladic figures, and provides insights into the treatment of prehistoric artefacts as objets d'art, and we therefore explore here its very considerable intellectual consequences.

We should first remind ourselves of the point and value of identifying the individual Masters, as that is expressed in the literature. J. Frel, introducing Getz-Preziosi's Getty essay of 1985, says: 'Instead of anonymous marble figures, we can see real people at creative work, communicating in spite of the passage of time and different ways of life.' Getz-Preziosi herself says that studying sculpture is thought 'to identify the personal styles of many individual, if anonymous, sculptors and, in some cases, to trace their development from novices to consummate carvers'. And the Virginia exhibition was perceived as making possible a one-man show 'devoted to a particularly important sculptor'.

Identifying the masters of Cycladic art: premisses and consequences of the method

The exemplar for the identification of masters in ancient Greece of course is the study of painted pottery, in particular black- and red-figure, as perfected by Sir John Beazley and his followers. Beazley was famously reticent as to his methods, though confident his eye could distinguish fine grades of distance between artistic personalities and saying 'that "manner", imitation, following, workshop, school, circle, group, influence, kinship are not, in my, vocabulary, synonyms'. Some of the Beazley attributions have been challenged or revised, either by Beazley himself or subsequently. In fact, since we do not know fully what criteria Beazley used, one cannot apply the same criteria fairly in attempting re-attribution. For an authoritative statement, we therefore go back to the original field where this Morelli method was perfected, the identification of masters in Renaissance painters, and in particular to the classic statement in Berenson's essay, 'Rudiments of connoisseurship'.

The first starting point is an identified corpus to gauge the range of a known artistic personality in the material under study:

'To isolate the characteristics of an artist, we take all his works of undoubted authenticity, and we proceed to discover those traits that invariably occur in them, but not in the works of other masters.'

This first condition is violated by the Cycladic corpus, for we have, and can have, no independently identified sets of works of 'undoubted authenticity' by any one Master, of the kind that is provided for Renaissance painters by signed, identified and documented pictures, and for Greek vase-painters by signed pots.

The next step is to isolate those little details of form which may unconsciously be characteristic of the hand of an individual painter – the manner of the ears, the hands, the folds of drapery, or the landscape behind the figure. The most revealing of all elements in Renaissance art is the manner of painting an ear: because part of the head, the ear is invariably painted by the master, rather than assistants, yet it is inconspicuous, unremarked by painter or by patron. This makes it 'more characteristic, indeed, than any other detail of the human figure'. Berenson says:

'the Italian painter kept on through a lifetime painting the same ear, because there was absolutely no call for changing it.'

What are the characteristic details in Cycladic figures that are equivalent to Renaissance ears? There are none! They do not offer significant details at all, because the stylized and simple form of Cycladic does not provide an aspect to the figures of that nature.

Usefully, Berenson, as well as providing an account of what is reliable for identifying the individual, also plainly states what it not to be trusted:

'The chin, the jaws, the neck, are all too typical, too easily copied, to be a ready indication of precise authorship . . . the followers and copyists of these masters share the same peculiarities, which can therefore serve only as indications of school, and not of the individual artist.'

The least applicable, then, are these larger features, those that define 'structure and movement in the human figure'.

Notice also that both Berenson and Beazley addressed painted objects, and – in the particular – painting styles which present much detail of form and technique in the visible particulars. Many of the Cycladic figures were painted, but the sculpted forms with which Getz-Preziosi has worked offers no kind of small-scale variation of the kind the method requires. It might be that a sculptural equivalent of the significant painted detail does exist – a recent study of Viking rune-stones, for example, hopes to distinguish individual carvers by the 'signature' left in the cut profile by a carver's tool and his manner of working with it. – but no attempt has been made to show that such a signature exists for Cycladic.

In summary, it seems to us that the identification of the Cycladic Masters rests on a mis-understanding of what the Morelli method consists of in any of its established uses. No significant details have been used. Instead, the larger features of morphology have been used to identify groups of figures that are, if anything at all, the Cycladic equivalent of Berenson's schools.

The identification of Masters, and the planning out of their artistic biographies, brings with it two further difficulties.

A first, noticed by Beard in respect of Beazley's vase-painters, applies with greater force to Cycladic. As we have no information about vase-painters other than what we see in the paintings, any discussion of the personality of the artists is, necessarily, a discussion of the paintings themselves conducted by proxy: 'There is nothing to be said about them that cannot be said about the pots themselves.' This is even more true of Cycladic figures, which are less forthcoming in what they seem to say to us. Beazley was again wisely reticent on this matter; he ends his essay on the Berlin Painter, written after fifty-five years of enjoying his 'friendly presence', by remarking 'Perhaps I ought to conclude with a characterization of the artist,' hesitates, and makes no attempt.

The second difficulty is whether the frame of Cycladic society, as that is understood from a broad range of archaeological sources, really provides for artists, whether as individuals, schools or circles, to develop their artistic personalities, in the manner of the full-time artist, replete with individual genius and a fiery creative imagination, as our era imagines him to be. The Cycladic islands – though falling between mainland Greece and the Cretan islands, and in later prehistory therefore placed between two regions of early states – have been economically marginal, with small populations. Broodbank has shown that the little 'frying-pans', identified as depicting ships, correspond to boats with a crew of some 25 paddlers. Yet even that complement is so much larger than the population of a prehistoric Cycladic village or hamlet that the boats must have been shared between several villages which could, together, provide sufficient to man the vessel; a ship from Melos would command as its crew one-half to one-third of the male labour force of the island. This is a society of subsistence farmers and fishers which has very little in common with those of the ancient world, in Egypt or among the Greek city-states, where there developed a social role for the maker of fine objects which might bear comparison with the artist's position in later European societies. A model for the place of art and of the artist in society that derives from the prosperous city-states of classical Greece or Renaissance Italy is not appropriate.

Identifying canonic proportions in Cycladic figures

The matching innovation in the study of Cycladic sculptures, to go with the identification of the Masters, is the identifying of a geometric canon in their dimensions. The search for a prehistoric canon of course follows from the canonic ideal as a known element in the classical Greek aesthetics of the later period. The figures are said to be planned by the division of their height into a whole number of equal units, usually 3, 4 or 5; the more complex figures, the harp-players, are said to be planned on a rectilinear grid of units; and the angles of the figure follow a harmonic principle of the 'golden triangle'. This proposition, again due to Getz-Preziosi, has been endorsed by other Cycladic specialists. Renfrew notes 'Dr Preziosi's important discovery . . . that canons of proportion were used in the setting out of the figurines themselves. This is clearly fundamental to our understanding of Cycladic art'. Getz-Preziosi explains how the formula worked, in her view. In the archaic phase, the sculptor conceived the human form as falling in three equal parts – one for head and neck, one for torso, and one for the legs. Arcs drawn with a primitive compass ensured that certain key points of shoulder and hip fell in with the formula. Later a four-part division was preferred, marked either with a curve or with a straight line, and a few sculptors experimented with three- and five-part designs.

These propositions need to be examined with respect to two aspects. First, do the dimensions of the figures actually follow these canonic divisions and harmonic proportions? A second is as essential: if they do seem to us to accord with a canon and a harmony, then – does it follow – were they designed by artists working to those principles?

Figures 1(a) and 1(b) reproduce Figures 17a and 17b of Getz-Preziosi's Sculptors of the Cyclades. They are identified as late classical figures of the Dokathismata variety, whose length is divided into four equal parts. In (a), a compass is used, working to a radius of one-quarter the length; and in (b) a straight-edge held perpendicular to the axis of the figure, again dividing the figure into 4.

Let us examine the divisions on each figure, starting at the feet. The lower division, marked 3 on Figure 1(b), falls about the middle of the legs in both figures; as the legs are drawn as even tapering curves, without distinct knees or other intermediate features, so division 3 is not a significant point in the composition. The middle division 2 falls, in Figure 1(a), well above the pubic triangle, just below the folded arms. In Figure 1(b), it falls just above the public triangle, well below the folded arms. In Getz-Preziosi's Figures 18a and 18b, printed in Sculptures alongside 17a and 17b, and reproduced here as Figures 2(a) and 2(b), the corresponding canonic division is set, in the case of 2(a) on the line of the folded arms, and in the case of 2(b), below the belt. The canonic division in the middle part of the body, then, is not a distinct point nor does it mark a distinct edge or feature in the composition; it is simply some level or other in the middle zone of the figure. In Figures 1(c) and 1(d), therefore, we strip out the lower division 3 as being a marker of nothing, and the middle division 2 as being a marker so inexact as to be worthless. The canonic proportions survive, but only as a 1:4 ratio between the head-shoulder portion of the figure, and the rest.

These Figures 1(c) and 1(d), bear only the top division-line, marked 1. In each figure it falls neatly at the top of the shoulders, at the figure's widest point. In (a), it is scribed with a circle, in (b) it is marked by a straight-edge. Notice that this choice of circle or line for the dividing convention materially affects the level of the outer points of the shoulder. In Figure 1(e) we present Figure 1(a) with a straight-line construction in place of the curve, holding to the rule that the line should run across the shoulders; its canonic proportion then becomes a 1:3.3 ratio between head and whole. In Figure 1(f) we present Figure 1(b) with a curve construction in place of the straight-line, holding this time to the rule that a 1:4 canonic proportion should be observed; in consequence, its uppermost dividing line no longer falls at the shoulder, and matches no other feature of the upper body.

We notice that the curving division line of division 1 in Figure 1(e) falls happily along the profile of the shoulders, where a straight-line construction could mark only the outer points. We notice that the curving division line of division 2 in Figure 1(a) falls awkwardly across the folded arms and coincides neatly with none of their vertices. In Figure 1(g) we therefore present a division that uses a curve for division 1 and a straight line for division 2. For division 3, we provide both curve and straight-line, interchangeable as this marks nothing in the figure. In Figure 1(h), we provide a straight-line for division 1, and a curve for division 2, and both again for division 3. Getz-Preziosi's canonic analyses use only curves or only straight lines in analysing any one figure; if individual sculptors are permitted to use one or other system in different figures, why are they forbidden to use both in the same figure? That flexibility in allowing straight–curve hybrids within the canon of 3 or of 4, makes them fit more varied shapes.

Figure 3(a) reproduces Figure 21a of Getz-Preziosi's Sculptors of the Cyclades a harp-player said to be from Amorgos. Getz-Preziosi identifies it as being made on an even grid of 8 by 6 squares. A certain number of points and lines in the composition indeed are seen to fall on or not far from the intersections of grid squares. Figure 3(b) reproduces the same figure, but the even grid is of 7 by 5 squares. Fewer points and lines seem to fall just on or not far from the intersections of grid squares than in Figure 3B, but this grid provides 35 rather than 48 points of intersection, so that does not surprise. Figure 3(c) reproduces the figure with a different grid, which like Figure 3 (a) has 8 divisions on one side and 6 divisions on the others; instead of being of an equal length as in Figures 3(b) and 3(c), they are of haphazard different lengths; again a certain number of points and lines in the composition fall on or not far from the intersections of grid squares.

Comparing Figures 3(a), 3(b) and 3(c), one finds some degree of fit in each case. It does not seem to us that the fit of Figure 3(a) is so decisively better than that of Figure 3(b) as to show the figure was made to a grid of 8x6 rather than 7x5. It does not seem to us that the fit of Figure 3(a) is so decisively better than that of Figure 3(c) as to show the figure was made to a grid of regular rather than irregular divisions.

An equivalent analysis can be made of the angles incorporated into the figures, which Getz-Preziosi sees as following the later aesthetic principle of the 'golden section'. Certainly golden angles can be found in some figures, but no attempt has been made to show they occur so often and so exactly as to prove that the figures were made according to those angles.

Various researchers have attempted to recover units of length or geometrical regularities of composition from assemblages of prehistoric or protohistoric objects. The common method is the one followed by Getz-Preziosi, to explore a body of material and to see what repetitive multiple of a unit-length, or what repeated geometries, declare themselves. The most through studies, and the ones most explored for the mathematical basis of the method, are Alexander Thom's studies of stone rings and other megalithic monuments in northwest Europe. Thom's scheme allows for the rings to have one of several different geometries. The goodness of fit varies; a certain number follow one or other variant in the scheme closely, some deviate by a little, some fit poorly. These discrepancies can be accounted for by the vagaries of constructing with large stones, by their being masses of a certain size rather than exact points, and by the many vicissitudes of subsequent damage, subsidence and human interference that might shift one or several stones out of their ideal place. Further, some rings which fit no standard geometry are shown to have a separate order of their own. In its essentials of method, Getz-Preziosi has taken an equivalent approach to Cycladic figures; in a body of material which presents the appearance of a geometric order, there has been found a small variety of canons, which in this case can be applied in two ways (curves and straight-line constructions), to which some figures fit closely, some fit less well, and some do not fit at all. Some of the exceptions, like the harp-players, can be accommodated by geometries of their own. The common lack-of-fit can be accounted for in a reasonable manner. About half of the classical four-part female figures are found to follow the canon, and about half do not. Some of these, Getz-Preziosi argues, 'were probably laid out quite carefully according to the canon, but as finished works carved in differing planes they no longer reflect the precision of the preliminary drawing made on the raw flat slab'. Other works show 'a conscious application of the canon but with modifications'; these 'departures from the norm were probably made deliberately' by an experienced sculptor, not a 'servant of the formula', who could 'adapt the formula to his own aesthetic sensibilities'.

Quite another interpretation of the geometrical regularities in stone rings is possible. Suppose their builders did not intend exact geometric shapes at all, but simply laid them out by eye into rough circles? In a most convincing study, Barnatt & Herring simulated the laying-out of stone rings by placing markers in a circle they judged by eye. They found it was hard to judge an exact circle visually, and the plans of rings that were arrived at in this simple way turned out to contain those geometrical regularities which Thom had thought to arise from a conscious mathematical canon. Two hypotheses therefore exist in respect of the regularities in the stone rings: either they arise from variability within some craft tradition in which placing is judged by eye; or they arise from the application of a body of complex and formal mathematics. We prefer the former view, by Occam's razor as the simpler explanation of the observed phenomena, and because it meshes better with our understanding of the nature of society and of knowledge in later prehistoric northwest Europe. We should prefer the latter view only if it is shown that:

We believe that a similar and simpler explanation exists for the geometries of the Cycladic figures than Getz-Preziosi's theory of the canon. In their common proportions, the figures do follow the proportions of the human body, as that has often been adapted in craft and art traditions. Her archaic three-part canon gives equal space to neck-and-head as to legs, as in other drawing and carving traditions which also reflect in their preferred geometries the greater importance of head and face than legs in encounters between human being. We notice the flexibility of fit, as illustrated in our Figures 1–3, which take as their worked specimens the exemplars of canonic geometry in Getz-Preziosi's main publication on the subject; we notice that if one permits divisions by different units, if one permits use of curved and of straight-line geometries, if one permits figures to deviate to quite some degree from the canonic ideal, and if one permits half the population of figures not visibly to follow the canon at all, then the canonic ideal becomes very flexible – dangerously so. Getz-Preziosi and others who have searched for – and found – the Cycladic canon have also permitted themselves parallel groups of non-canonical figures which do not have to follow even these flexible rules. It has not been shown that:

Following Occam's razor, therefore, and because the abstract principles of academic aesthetics embodied in the proposed canon do not mesh well with other evidence for the nature of Cycladic society, we prefer the simpler idea, that these canons of proportion did not exist in the minds or the hands of the carvers of the Cycladic figures. The same will go for the angles of the golden section.

A good instance of canonic principles displayed in a body of ancient art is Egyptian, especially in its earlier periods to the high Old Kingdom, or to about 2100 BC – much the same period as Cycladic. But see how varied is the iconographic evidence itself, how exactly it follows canonical rules, and see how many and how compelling are the circumstantial sources of evidence that the highly evolved and structured Egyptian civilization was inclined to these ordering principles. Yet even there, as Davis' new study of the Egyptian canon notes at its very beginning, the invariance of Egyptian art 'cannot simply be taken for granted'. It simply will not do to assert, without well-founded proof, that Cycladic art follows a canon because a few chosen figures seem to fall near one. Yet a growing literature, as recently Renfrew's Cycladic spirit, discusses the supposed canon without the necessary first step of showing that the canon exists; until that has been done – and our preliminary investigation leads us to think it cannot be done – we disagree with Renfrew's commending the canon as 'a line for further investigation'.

Birth of the genius that was Greece

One reason that Cycladic figures appeal to modern aesthetic taste is that they are perceived as being in some way precursors to western art. Thimme has commented: 'Today, four a half thousand years later, they enchant us again with the purity of their forms, their freshness, and their perfection. For us they radiate vigor and originality – and they mark the beginnings of European art.' For Colin Renfrew 'these Cycladic marble ma idens mark the birth of western art. For we see in them some of those same qualities which so delight us in the korai of the Acropolis'. Or again, 'In the art of the Cyclades there is something fresh and arresting, as we see the first steps being taken towards the creation of a European civilization'. Yet it has not been shown that these Bronze Age figures from the islands are directly linked to the continuous tradition which indeed links the kouros sculpture of the island and mainland Greek Archaic, a millennium later, right up to the sculpture of the late twentieth century. No continuous sculptural tradition from the Early into the Late Cycladic period has been shown, let alone one from Late Cycladic into the archaic and classical periods. Cycladic figures were known to some individuals in classical Greece, for a certain number have been found in archaeological contexts of the classical period, but there is no evidence to suggest these in any way influenced the art of the later period. One figure was later built into the wall of a Turkish-period house at Grotta on Naxos: is that to be taken to make Cycladic the birth of Ottoman art?

It is essential here to distinguish a historical succession, a chain of connections by which one aesthetic is transformed into another, from a chance precursor which happens to pre-figure a later, or to echo an earlier, aesthetic. For an example of the former, one has only to look at the classical sequence in Greece or the aesthetic chain from Greece to Rome and then to Renaissance Europe, and to think of the way the Renaissance was a conscious re-birth of ancient aesthetics. But a simple coincidence of place that involves cultural traditions at distinct and distant periods does not of itself amount to a continuous tradition. No evidence links the Cycladic carvings of the Bronze Age to, even, the Archaic of the Greek mainland some 15–20 centuries later. The chain of classical Greek sculpture goes back continuously to the kouros and beyond that, in a disputed degree, to sculpture in Egypt and elsewhere outside Greece. The aesthetic of the kouros and of the Cycladic figures are both 'simple' or 'primitive', but they are different in all particulars, and they have no historical connection. So they are not the first steps towards the traditions of western art that grew in and from classical Greece. They are early steps in a direction which happened later to be followed by the traditions of western art. That, and only that, is why they seem familiar to us, why 'we see in them some of those same qualities which so delight us in the korai'.

Since there is no real historic continuity between Cycladic and the first, Archaic classical, the enthusiast who wishes to join them must instead assert a universal aesthetic, take an indirect route or prefer a more mystical argument. It is this view that these prehistoric figures indeed spoke to the sculpture of the archaic and classical periods of Greece and then somehow to us today that Joseph Masheck, Professor of Art History at Hofstra University, takes in his essay 'On Neolithic ultramodernity', He remarks, 'In them, it is sometimes as if the Greek archaic proper, that first dawn of classicism itself, were still just beneath the horizon.'

This desire to find the roots of western civilization extending into the fisher-farmer societies on the barren soils of some rocky islands in the middle of the Aegean is curious, when there are developed states and civilizations across the east Mediterranean – in Anatolia, the Near East, Egypt – with documented links to Greece. The diagnostic clue is, Masheck explains, in the values of marble: 'Marble: raw substance of entire mountains, sea-smoothed pebbles washed ashore. Marble catches the light and the imagination. It invites touching and handling. The bright, often translucent appearance and tactile properties of white marble are very compelling indeed.' Although the medium of marble might touch both the third millennium BC and the second millennium AD, this does not demonstrate that the roots of western civilization lie in Greece.

This attempt to privilege Greece is not uncommon. In the history of western art, Gombrich looks back to the 'Greek Revolution' with the introduction of narrative art. Yet many of the motifs used on painted pottery in Early Iron Age Greece were merely borrowed (and misunderstood) from the Near East. The Greek alphabet derives from the Levant coast. The introduction of monumental sculpture to Greece in the archaic period, notably the striding youths or kouroi has long been seen in the light of possible influence from Egypt. It is however the 'Greek genius' which has been emphasized: in democracy, in theatre, in philosophy. In this improving story it is easily forgotten that Greece also 'gave us' slavery and that Athens' ability to beautify itself was built on a consuming imperialism. It is the visible attempt to isolate Greece from its eastern (and certainly southern) links which has surfaced in the contemporary debate about Martin Bernal's Black Athena. The collector, like the Getty and like White and Levy, who wishes to extend backwards from a nucleus of classical material has a choice between a cultural continuity, which leads towards Egypt, and a geographical continuity, which leads towards Cycladic. The Getty, like White and Levy, has chosen Cycladic.

How is the cool historian of art, recognizing the chronological gap to be bridged in linking the Cycladic genius to the Classical? Martin Robertson takes 'art in Greece before the Iron Age' as the 'prologue' to his History of Greek art. He identifies the Cycladic marble figures as 'the first sustained production of objects of aesthetic interest in Aegean lands', and notices that the figuring of musicians provides evidence of two civilized crafts: 'The quality of these figures, and the evidence they give of musical attainment, suggest that quite a forward stage of culture had been reached in the Cyclades at this time.' Without a documented link, a bridge must be made in a slightly mystical manner from Cycladic to the Late Bronze Age, Mycenaean world: 'These are not cases of "influence", but perhaps of some kind of inherited slant.'

An aspect to this issue is the changing relative values given to European and to non-European cultures, as the modern age of empires come to an end. In the first half of this century, the British and French empires spanned, and Europe saw its cultural roots also to lie at a distance, in Childe's 'most ancient east'. It remains the case that the first fundamental of European later prehistory, the adoption of agricultural ways of life, came into Europe from Asia, although megalith-building and some aspects of metal-working – other imported skills in the old diffusionist view – are autonomous European inventions. The bundle of traits we call 'civilization' in the Greek city-states also derives in large measure from the East. These facts of history fit less comfortably the mood of the early 1990s, when the Soviet Union, last of the modern empires, has unbundled itself, and when Europe envisions its fine future as a strong confederacy of autonomous nations freely joining to work together, a Europe independent of other continental powers and self-sufficient in its spirit. Archaeological proofs of a united, autonomous and pan-European history are welcome. The strongest precedent is offered by the Celts, a distinctive and fine artistic tradition and ethnic identification which is pan-European yet distinctly of Europe alone, open to the world in its dealings with non-European cultures yet reserving its own European identity; a passion for Celtic matters is evident at present among the European exhibitions. The Cyclades make possible a similarly agreeable source to the European character, which can thereby be traced through the pan-European empire of Rome (another precursor of the modern European Community), back to classical Greece, and then – not to confused and confusing countries in the modern third world of Africa and Asia – but to Cycladic figures (both miniature and life-size) and beyond them to the Neolithic carved and ceramic figures of the Balkans and Greece; the whole story is safely, reassuringly contained within the boundaries of European civilization, in the lands of the European community as we wish to define it today. It is not a coincidence that the wish aesthetically to vault from classical back to Cycladic is strong.

Renfrew has taken particular note of the big Cycladic figures, the largest now approaching life-size. Noting the rarity in early artistic traditions of life-size human depictions, and its importance as indicator of a precocious and advanced aesthetic, he regards the scale of these eccentrically large examples as essential to Cycladic esteem. Setting the questions of provenance and authenticity aside, we instead agree with Broodbank, who comments:

'is it realistic to look upon the equally uncommon monumental figures as the outcome of heroic and precocious attempts at a life-size portrayal of the human form . . .? . . . what we really see is a scaling-up sequence from the smaller pieces, via intermediate examples of unusual but sub-human size. Is it therefore not more plausible to view the largest works as an outcome of status-driven competitive emulation – of a species of Peer Productivity Interaction? This is not to belittle the skill needed to make them, nor to deny that the by-product is indeed a life-size rendition; it is merely to distinguish between ancient intention and modern conception.'

Conceptions of Cycladic and historical scholarship: the case of the baitylos

The Cycladic industry is funded by the collectors, their singular passions, judgements and concerns. Naturally, many take an amateur interest in what they collect, and wish themselves to learn more. The museums and the market-makers are encouraged, even obliged, to take notice of these well-meant enthusiasms, and a perceptible respect is therefore offered to the collectors' ideas about their treasures, even where it bears little relation to any reliable historical scholarship. After all, a primary obligation on the museum curator is to acquire and, in the present mood, this means humouring and nursing the collector in the hope that a present expectation may be turned into a real acquisition. The point is illustrated by the identification of some Cycladic artefacts, those that seem to be simple shaped pebbles rather than naturalistic sculptures, as 'sacred stones', or baetyls. In following this religious interpretation of small pebbles in Early Cycladic graves, the Erlenmeyers and others have looked to classical cults of Asia Minor and Cyprus where sacred rocks were worshipped as deities. These rocks or baetyls were usually conical in shape, and their images are often found on the coinage of the city. One of the most famous stones was that at Perge in southern Turkey which was identified as the goddess of hunting, Artemis. Another was located at Paphos on Cyprus which was worshipped as the goddess of love Aphrodite. Elsewhere on Cyprus, near Amathus, an inscription indicates a cult place of 'the Seven within the Stelai', presumably a group of seven baetyls, perhaps also dedicated to Aphrodite. The presence of baetyls at several cities in the eastern Mediterranean during the classical period gives us no ground to interpret these polished marble pebbles – which are considerably smaller than baetyls – as sacred rocks so many centuries earlier.

This confusion – between very large stones in Anatolian temples of the Classical period with known function and meaning, and very small stones in Cycladic graves of the third millennium with unknown function and meaning – rests on the most elementary error of scholarship. By chance, it recapitulates a previous error of research, in which the idea of baetyls again was the source of confusion. Late in the 19th century, the senams, certain stone monuments of the north African littoral, were identified as baetyls, sacred stones of some kind. Large, rude, and erected in the form of doorways, they were also connected to the prehistoric megalithic structures of Europe, and specifically to Stonehenge in southern England, whose stone trilithons they seemed to match closely in their form. Decent archaeological fieldwork soon showed them to be ruins of Roman olive-presses, not baetyls, not prehistoric, not anything to do with Stonehenge. A century on, the collector's fancies drag into Cycladic studies an equivalent error – or a worse one, since at least the Libyan senams were much the same size and shape as Stonehenge trilithons.

The errors in the logic of the baetyl are clear. More subtle are the discrepancies in reasoning like this:

'Since I view the canonical figure as a portrayal of the Great Goddess of the Cyclades, and since her image was found in at least one case (on Keros) in a grave with two musician figures, it seems to me logical to suppose that the Cycladic harpists, flautists, and syrinx-players were closely connected with the goddess. Representations on Sumerian reliefs of the third millennium provide the closest parallels for the Aegean harp-form and repeatedly show harpists present at cult banquets. Anton Moortgat regards these scenes on votive reliefs as representing the Feast of Spring and the New Year. According to Marie-Louise and Hans Erlenmeyer, the reunion of the goddess with the god newly returned from the lower world was celebrated at the New Year's Feast. This celebration, as it is shown in the Sumerian representations, must have resounded with harp music. It does not seem to me too far-fetched to associate similar religious practices with the Cycladic musician figures and to view them as attendants of the Great Goddess who accompany her with music as she rises from the underworld and ascends to the heavens.'

Notice that the well-meaning and ill-informed concepts of the Erlenmeyer, amateurs in these matters but – importantly – wealthy patrons in the Cycladic industry, figure with a large respect in this speculation, which mashes together a variety of fragments (rather far removed in space, in time and in social context from Cycladic) and some fictions (like the nature of New Year's Day ritual), into an ancient vision that is based more on soul than on scholarship.

Again:

'The hypothesis that Cycladic idols serve the cult of the Great Goddess and Mother, the Queen of Heaven who descends to the underworld and from it ascends to new life, who possesses a daughter and is occasionally accompanied by musicians: such an interpretation, I believe, is in complete accord with all the different forms of the canonical idol known to date.'

Presentations and perceptions in Cycladic studies: the choice of words and the values they carry

It has recently been fashionable to emphasize the subjective element in how we view the past, the role of our values and preconceptions; in the more radical view, we construct a past in the present to suit ourselves. It is certainly true that we reconstruct the past within, necessarily, a modern world; we make sense of past insofar as it intersects with our modern perceptions. We express our perceptions about Cycladic in what we say about the objects, and in the way we look at them. Running through the newer literature on Cycladic is a vocabulary of fine art, and with that vocabulary is carried along some conceptions about Cycladic society as a fine society, a vision which is at odds with the archaeological picture of poor islands with small populations living at a simple subsistence level, without great towns or patrons of the visual arts. Many words in the literature can be explored. We indicate first the general point by offering a choice of sentences to describe the manufacture of a Cycladic figure, and then address three examples only, two in words and one in the physical: the word 'Master', the word 'civilization', and the habit of displaying Cycladic figures upright.

Consider these three short sentences:

'The Master sculpts the idol in his workshop.'

'The carver crafts the figure in his hut.'

'The rustic shapes the pebble in his hovel.'

Each sentence describes the same act in words that are near-synonyms in their first meanings, but they make the act high, middle or low in its social implications, in the level of artistic refinement and in the social and the physical context of labour. The placings of high, middle and low are themselves statements of attitude, setting craft as a respectable skill but a lesser gift than a fine art.

'Master'

Let us consider the word 'Master' in respect of a maker of Cycladic figures. The Oxford English Dictionary defines 'master' in this sense as 'An artist of distinguished skill, one of those who are regarded as models of excellence in their art', and the ideal of the master with his school or workshop, as the social framework where the making of fine things can best be accomplished, of course derives from the practice of medieval and Renaissance Europe, as Alberti and others describe it. So the word 'Master' – not quite the same as the word 'master', since the initial capital itself carries meaning – refers to more than accomplishment; it goes with the word 'masterpiece', the piece made expressly to use and to show off the Master's skills. And it also goes with the places of art, of the artist and of the connoisseur of art as these have been experienced in the western world these last few centuries. When the Morelli method, and a fuller knowledge of the historical authorities, made it possible to define un-named hands and personalities, distinctive in the manner of the known, named masters, they could therefore be defined as Masters, the 'Master of Flémalle' or the 'Master of the Life of Mary'. Beazley, applying the Morelli method to Classical Greek painted pottery, identified named and anonymous potters and painters, like the 'Berlin Painter', together with their concomitant satellites of 'group of', 'class of', 'circle of', 'manner of', following the medieval and Renaissance model again of 'school of', 'workshop of', 'manner of'; a parallel framework has been followed for the study of the Greek classical sculptor. In this way, the context of artistic production in later western societies has been projected back on to the classical world. Perhaps pots were made and painted in classical Greece in a context of masters, assistants, workshops and schools so closely resembling those of easel-painting, sculpture and decorative arts in postmedieval Europe that just the same words are in order; but that similarity needs to be demonstrated, rather than carried in as the intellectual hand-baggage of a convenient set of words.

In a further projection, the later western framework has now been projected back on to the Cycladic figures, with two telling differences. The first – which frankly puzzles us – is that the carvers are called Masters with a capital M. Although classical Greek painted pottery is one of the most technically accomplished of all ceramic traditions, and although Beazley had the highest regard for the great hands in classical pottery, he stopped short of the word Master in his later writings, preferring the more reticent 'Potter' and 'Painter'. We think he was right, for 'Master' carries with it certain implications of a social framework that the more neutral terms of potter and painter, and for Cycladic figures the word 'carver', do not bear. Getz-Preziosi, fully understanding how small the populations of the Cycladic islands may have been, and therefore how few and solitary may have been the carvers, nevertheless chooses the word 'Master', a word that is high rather than middle in its aesthetic and social implications.

Other projections from later aesthetic traditions in the same way have come to be applied to the Cycladic figures in the same way. Getz-Preziosi divides her sculptors into 'archaic', 'classical' and 'late classical' phases. Thimme follows a similar scheme, defined by a formal aesthetic progression from a Precanonical, through to the Canonical and then the Postcanonical. Precanonical figures lead to 'the establishment of the canonical type' including 'the hesitant beginnings of the folded-arms position'.

This series of projections depends for its safety on a series of premisses, which deserve notice.

A first is that the model of Master and workshop indeed fairly represents the conditions of under which works of art were made in early modern Europe. We here accept that it does, but also notice the intellectual context under which this ordering was made. Just as the nineteenth century saw the sorting-out and putting into order of the collected objects of archaeology, with the three-age chronology imposing system on prehistory and the essential frameworks of classical archaeology being built, so the same period saw an equivalent chronicling, cataloguing and ordering of the subjects of the art historian's study. There the essential framework was authenticity, so as to define the central corpus of the named masters, and the penumbra of other and lesser works, 'school of', 'copy of', 'after', 'in the manner of' that surround it. Now, this intellectual need, to bring order into the multitude, found a corresponding need to establish an equivalent order in the artistic market-place, to sort out which of the countless paintings 'said to be by Rubens' that had been carried proudly home from the Grand Tour really were works of the Master, and which were simply 'bought as Rubens'. Out of it came the clear hierarchy of quality, and correspondingly of price – a hierarchy that goes from the summit of a certain work by the Master to the trough of a copy made by anonymous hand at some subsequent time and in some distant place – which structures the art market to this day. It is expressed for example in the variety of phrases in customary use in the auction catalogues that run from the artist's full name, indicative of a secure attribution to the Master's hand alone, down to 'manner of' for the least authentic. The orderly running of the market enjoys this familiar structure, which is duly emerging for Cycladic figures. As well as the named 'Masters' of Cycladic art (whose reliable identification we addressed above, pages $$), the attributions are now being extended to the followers of the Masters: a 1991 Christie's sale catalogue identifies one figurine as 'in the style of the Schuster Master', duly beginning to extend out to Cycladic the sale-room structure of art. The concept of the Master is itself embedded in the larger scheme, of an artistic cycle of birth, florescence and decline, which is shared in common with the archaeological expectation of three-phase early, middle and late (or pre-Classic, Classic and post-Classic) structures.

Both the scholar and the market have a need for a secure order, and a hierarchy of values, but there are two dangers in that coincidence of interest.

The more obvious danger is a direct corruption. Tragedy came to Bernard Berenson, the greatest Master of the Morelli method in its first application to European painting, when he allowed the two interests to be fused together, the market interest corrupting, and over-riding, the scholarly interest. He became associated directly with transactions in the market, so that his scholarly eye was turned also by a personal market interest, and his intellectual ruin followed. No such public calamity has been seen to occur in the study of classical pottery, fortunately, and no public scandal has followed from the many close associations of consultancy, identification, and writing of sale catalogues between scholars and market-makers in the Classical field, or in Cycladic. Yet there is a real conflict. Distinguished scholars identify, authenticate and explore the items in – to name one of many – the White and Levy collections; great museums celebrate them in the finest of public places. Auction-houses and dealers .

It is significant, we think, that the senior classical archaeologists who direct the Beazley Archive in the University of Oxford, a world-resource for attribution of Classical ceramics, in 1991 made public their concern as to how its resources are used:

'The Beazley Archive has been increasingly concerned about protecting its reputation in the scholarly world against appearing to condone the trade in illegally exported antiquities, by allowing the use of its resources for the enhancement of the value of Greek vases of no clear origin.'

Accordingly, the Archive has resolved:

'Our intention is to ensure that the Archive is not, and is not seen to be, used in any way which might lead to the commercial enhancement of objects which may be suspected of having been removed illegally from the country in which they were found.'

Also in Oxford, the University Research Laboratory for Archaeology and the History of Art has decided not to perform thermoluminescent testing to authenticate terracotta figures from west Africa, specifically Mali, which are being looted and smuggled for the market in much the same way as Cycladic.

We do not suggest in any way there is any direct scholarly corruption in respect of Cycladic figurines. But we do think that the co-existence of the different interests is becoming more difficult. In Cycladic, as in other fields, one notices the same experts are active in scholarship, in the museums, and in advising for the market. And one notices how many of the attributions include and depend on 'probably', 'possibly', 'perhaps', and the insidious 'said to be'. 'Said to be' – by whom, to whom, with what motive, on what authority? And how often may 'said to be' stand for 'wannabe'?

The less obvious, and the more real, danger is the treating of the two interests, those of the present market and those of the historian of the past, as if they were wholly equivalent. There is no harm in building a structure for the Cycladic market entirely to suit the market's convenience, with its greater and lesser Masters, its enigmatic Masterpieces that have yet to find an identified Master, its school of this Master, and its style of that, provided it is also and always understood that the whole masterly structure may be a modern public convenience, not a re-creation of the social and intellectual realities of the prehistoric Cyclades. Yet, in a real world in which the scholars and the market-makers live closely together, the convenience of one comes to be imposed on the other.

The earliest classification of figurines named its identified types after the sites where examples were found, the Plastiras type for a cemetery on Paros where four such figurines were found in a single grave, the Louros type for a grave at Louros Athalassou on Naxos where seven of the figures had been placed in a niche. From a scheme which named the figures into an archaeological context, connoisseurship has preferred a framework which is linked to possession.

The use of the term 'Master' by Getz-Preziosi in itself might suggest that these figures were created by a single man. Her sixteen 'sculptors' are named in two ways. The two in the first group are named after excavators in the Cyclades: C. Doumas and N.M. Kontoleon. The fourteen in the second group are named after the present or previous proprietor of type-figures, either museums and private collectors. The original masters, those of Renaissance painters, are generally named after the subjects of defining pieces, like the Master of S. Giles for the painting of the Legend of S. Giles, or the place where the defining piece resides, like the Master of Moulins. The new habit for Cycladic follows the place or the present private proprietor: we have a Steiner Master as well as a Naxos Museum Master, but we have no Master of the Folded Arms, or Master of the Biggest Ears. This is a small matter, but one which tells of the modern dominance of the proprietor over the character of the thing he possesses.

'Civilization'

A division between the civilized – 'we' – and the barbarians – 'they' – was central to the classical Greek and classical Roman views of their world, and the higher ideals of their own societies. That division, and attitude, remains with us: we are proud of our civilized societies, but conscious the barbarian forces may at any time mass at the gate to pillage and burn us out. Barbarian remains a term of denigration. The archaeological definition of civilization is in other terms, by reference to a number of definable traits, the traits that go with cities and with state societies – a large population of 20,000 plus, a class-based hierarchy under king or emperor, an army, centralized bureaucracy, tribute-based taxation, established law, urban settlement, frontier defence, roads, a priestly class, pantheistic or monotheistic religions, palaces, temples and other public buildings. These primary traits are, for the most part, evident in classical Greece. All are absent from the Bronze Age Cycladic islands. Cycladic was a barbarian society, not a civilization. Davis remarks, of the inappropriate term 'Cycladic civilization':

'Life in the islands in the Early Cycladic period bore none of the hallmarks of civilizations: no, or few, large communities. no written records, no spectacular buildings, and no palaces of the sort built later on Crete and at Mycenae to house organized central governments. On the contrary, early Cycladic settlements were small, the buildings modest – of small stones set without mortar – and graves were seldom richly furnished. (Modern visitors to an excavated settlement or cemetery would find little to hold their attention.) It is likely that all or most settlements were independent, and no evidence points to the political unification of an entire island, let alone of the Cyclades as a whole.'

Of course, if one follows the woolly notions of space and time that are evident in dreaming of the baetylos in Cycladic prehistory, or of Cycladic peasants following Sumerian state theology, one may look to a society a little removed in space, whether on Crete or on the Greek mainland, and more than a little removed in time, and – finding it agreeably civilized – choose to regard Cycladic as pretty well civilized, which by any proper historical scholarship, it is not.

Cycladic figures: lying or standing?

We linked, above page 000, the rise of regard for Cycladic to the convenience of these objects in an era of smaller dwellings. We see Cycladic figures as having the presence of large objects, yet they are small enough for the mantelpiece: 'These Cycladic figures, though small in size, have a monumental, hieratic presence.' On the Manhattan mantelpiece, or in the museum display-case, Cycladic figures are invariably placed upright. Yet they do not stand upright unsupported, and are now always held in the vertical by mounts or by wires round their necks. Clearly, this is how we choose to see Cycladic figures today, and the vertical rule derives from the upright figure as the common form of the large, grandiose, heavyweight sculpture, whether one of a thousand Lenins leading the masses to a proletarian victory, or a Winston Churchill in Parliament Square. The monumental miniature follows the upright model of the full-scale monumental. Whoever saw a massive Lenin lying down, except in disgrace after the fall of the Communist Party monolith?

What about the way the figures were held in prehistory, whether in life or when placed in the grave? For the reasons exhaustively discussed above, the direct evidence is slight. One can begin with the observation that figures do not stand safely on their own feet, nor are the feet set in or on a base. Yet the feet of the figures are carved with care and detail, with individual toes even on small figures. It therefore appears that most figures were made neither to stand by themselves, nor to be inserted into stands or pierced holes that would support them upright. Getz-Preziosi using what evidence there is, envisages the figures as being held or carried in a horizontal position, small figures in the hand, larger figures on the two out-stretched palms of the bearer. In the North American catalogue, Getz-Preziosi says of their presentation, 'One or two of the characteristic folded-arm figures are shown in the reclining position intended by their sculptors . . . but, for reasons of visibility . . ., the majority of the works in the exhibition are displayed and illustrated in the conventional manner.' Getz-Preziosi again: 'in the sterile environment of a display case, the figure looks stiff and fragile, unnaturally – and, I dare say, incorrectly – posed in the tips of its toes'.

The vertical pose of the Cycladic figure today, then, arises from modern habits in monumental sculpture rather than such clues to prehistoric attitude as we have. And because Cycladic figures will not stand up on their own, they are now inserted into moulded perspex blocks to make them upright: the physical aspects of their presentation begin to follow the intellectual consequences.

Cycladic figures: white or coloured?

One of the best-known instances in all archaeology of a discrepancy between ancient reality and modern perceptions concerns the colour of Classical marbles. Originally painted, they lost their colour in their time in the ground, so that classical revivals have, over the centuries, revered the surface qualities of clean, austere white marble rather than the painted ancient ideal. Even today, when one has knowledge of that transformation, the sight of a Classical figure painted into its original bright finery shocks; it is so different, so alien, so removed from what one has become used to.

Precisely the same story applies to the Cycladic figures. Many bear traces of blue and red paint, or 'ghost' areas where the paint has left a different surface texture. Renfrew says of our high regard for the simple power of the whitened figures:

'We see Cycladic figures as pure in their whiteness. The absence of surface detail emphasizes the simplicity of form and all is reduced to clean lines and the polished surface of Cycladic marble. but it was not always so.'

Broodbank comments:

'Let me suggest a visual experiment. Trace a figurine on to white paper. Add (not in time-faded hues or modest black-and-white stippling, but bright hematite red and azurite blue): head and pubic hair, lycanthropic eyebrows and staring almond eyes, densely tattooed dots or vertical gashes over most of the face, and necklaces and bracelets at throat and wrist . Alternatively, look at the rare figurines that do depict features such as eyes, ears or baldric in relief; these get a hard time for their "fussiness", but in fact tell us much about how their bleached sisters once looked. Whichever method one chooses, what looks back at us is something extraordinary but utterly "other". Detail in no sense intends to emphasize form or line, and, if anything, distracts from it. Rather, its aim is to send out messages significant to an early Cycladic observer. Indeed, the range of tattooed motifs and hair arrangements attested may indicate that the painted surface records crucial messages concerning diversity over the neutral and standard canvas of the marble form.'

Broodbank notices that other Cycladic artefacts, such as the vessels densely covered with elaborate painted or stamped and incised designs among the prestige pottery of the Keros-Syros culture, suggest that great importance was attached to surface detail. If the Cycladic figures were commonly painted, and if their simple beauty is unusual in Cycladic material culture, the loss of paint may be the central issue in understanding them. The simple forms, and the lack of detail, may arise only from their being taken back, by the chance of time, into an unfinished state: what we see and delight in are the plain, preliminary blanks which were made primarily in order to carry the essential variability of coloured decoration – the meaning that mattered. If that were the case, then the whole modern apparatus of regard for Cycladic simplicity is in error. Instead of the supposed Cycladic Masters having arrived at the same aesthetic as Modern Masters, it simply chances that unfinished Cycladic figures chance to resemble finished 20th-century sculptor.

Modern perceptions of ancient Cycladic society

We have looked behind the words 'Master' and 'civilization', and at the upright attitude that Cycladic figures are now required to adopt. Many other words that run through the newer Cycladic literature – enigmatic, inscrutable, hoard, idol, sculptor, art – will on examination be found to bring with them a variety of meanings and nuances which may not be applicable to Cycladic figures in their prehistoric meaning.

Since all words hold meaning beyond their dictionary definition, anything whatever that is said about Cycladic figures (or anything else!) will bear nuances. To speak neutrally of them is to be silent. Yet it is inescapable that the vocabulary of Cycladic together provides a certain atmosphere, perhaps already evident from the quotations in this paper, and particularly visible – to our reading – in Getz-Preziosi's Sculptors of the Cyclades. Here is a sample of the atmosphere created by the style, when some questions of chronology are addressed:

'While the earlier part of the second phase was a time of exuberant self-confidence and virtuosity in Cycladic sculpture, toward the end of the period the spirit of the times seems to have changed, to judge by the radical differences of iconography and style now seen in the sculpture.'

The vocabulary sets the frame of thought. It predisposes one to the idea that Cycladic figures might have been idols; it predisposes one against other interpretations for which there is equal evidence.

Sometimes, the frame of thinking is more explicit, and seems more directly to steer thought in a certain way. Getz-Preziosi sub-titles her study of the sculptors, 'Individual and tradition in the third millennium B.C.', and elsewhere talks of the Cycladic figures as 'orphans': 'With orphaned Cycladic images, as with orphaned images from other regions, one can never be certain in what context they were used, although the chances are good that it was a sepulchral one.' Is there not to be heard here an echo of that common choice of the word 'individual' in contemporary north American society, a different word carrying different nuances, where in other times and places the words 'person' or 'people' might have been used? In contemporary western culture, an orphan is a little soul, lost through misfortune, that cries out to be adopted and given a home; a good home will compensate for what it has already suffered. Does not the word 'orphan' carry a similar burden, behind it the fear of losing self, losing the individual, becoming nameless and lost? In these ways we take the figures from their makers in prehistory and bring them into the values and into the psychologies of our own world-views. To note the proprietorship of a figure as being in the 'Mr & Mrs Stanley Marcus Collection of Fertility Figures' at the Dallas Museum of Art is to make a statement of their ancient and their modern meanings.

Art, aesthetics and human universals

A recurrent element in the evocative strain of Cycladic literature is an appeal to timeless universals: the figures speak directly to us across the centuries because they share an eternal language for all humanity: 'With minimal means, they attain an intensity of abstract expression and subtlety of form that, though totally idiosyncratic, have universal appeal.' This encourages a rhetoric that leaps from islets of the prehistoric Aegean to societies of a very different character. Here is that jump in the annual report of a famous Americanmuseum:

'The fall began auspiciously when John Coney's monumental silver punch bowl, acquired through the Theodora Wilbour Fund in memory of Charlotte Beebe Wilbour, was placed as a centerpiece of the special exhibition of American silver. Coney (1655–1722) was a Boston craftsman who brought Georgian British standards of work in the precious metals to the Colonies. His bowl is the forerunner in size and simplicity of what Paul Revere was to create decades later in the eighteenth century for the Sons of Liberty in Massachusetts. The Coney bowl's shape compares provocatively with two marble vases of about 2400 BC, found years ago with two Cycladic idols of elementary form in an early Bronze Age tomb at Marathon on the eastern coast of Attica.'

Why do these things compare? why is the comparison thought provocative? First, one must take full account of the essentials of three-dimensional geometry. It is a restricted system; it permits only a certain range and variety of forms. Second, even that range of forms is vastly reduced, for practical purposes, by the physical constraints on the made form of the artefact that follow from simple requirements of function, by the physical properties of the materials of which it is made, and by the functional requirements of a human user. A bowl or vase made of silver or of marble that is to work well is likely to be held within a narrow vocabulary of forms. It does not follow that there is any close link of character or of spirit between the maker of the one and the maker of the other. A similarity of form does not make the Israel Museum Master into the twin of Brancusi. And because time has direction, the similarity of the Cycladic to the Brancusi is not to be equated with the similarity of the Brancusi to the Cycladic.

Esteem for Cycladic is one of many examples of a broadening of aesthetic tolerances among the connoisseurs of this century. With that broadening of aesthetic tolerances there needs to go a matching recognition that there is no simple and close coupling between art of a first-class quality, as we choose now to perceive it, and the virtue of the society that made it. The maker of a superb contemporary piece may be a violent, deranged, horrid, absurd person. The society of the prehistoric Cycladic figure-makers was the world of subsistence fisher-farmers on marginal little islands; they were none of them people like us.

Time has now chosen to give a curious twist to these ideas of timelessness. The 'Modern movement' of the early decades of the century itself has become a historical phenomenon. Modernist buildings have become period pieces, the masters of Modernist architecture are out of fashion, and a post-Modern architecture follows other values. The same goes for the sculptures of Brancusi and Modigliani and Moore, although the critical view of Modernist sculpture is kinder in the 1990s than it is to Modernist architecture. Just as the comparison of Cycladic with the Modern movement has become more compelling, so has regard for the Modern movement become less compelling; no longer where the best of artists are, is just some point where some artists once were. While Modernism respected the eternal verities created by the rule of a form that followed function, Post-modernism in the 1980s took to arbitrary games of pastiche and play, the breaking of rules, the mixing of different orders, the reviving of any historical style that takes the fancy. The Modernist aesthetic of Getz-Preziosi, Renfrew and Thimme has set aside the eccentric, anomalous and unconforming figures; it has corralled the whitened Cycladic figures that it prefers to revere within a set of austere rules and placed them within a narrow historical trajectory of forms considered acceptable at a given stage of their forward evolution. A post-Modernist view of Cycladica rejoices instead in the garish colours, regrets the clean neatness of the scrubbed white surfaces, celebrates the quirky and the unusual, delights in the breaking of the canonic rules (if such rules existed) – especially if done with a mischievous purpose – and enjoys the knowledge that some of them are fakes.

Connoisseurship and archaeology: an essential asymmetry?

The purpose of this paper is simply to document the consequences, material and intellectual, of a recent rising regard for Cycladic figures as objects for the connoisseur's zeal. In closing, however, it is right to identify a fundamental asymmetry in the consequences for the object, and for our knowledge of it, between the archaeological and the connoisseur's view.

For the connoisseur, the value of the object largely resides in the thing itself; a knowledge of context is a useful accessory to that knowledge, in telling of its background and in reassurance that the thing is not a fake. Nothing fundamentally differs between a Cycladic figure recovered in a controlled archaeological excavation and a Cycladic figure that 'surfaces' in Switzerland.

For the archaeological viewpoint, the information immanent in the object itself are elements of a larger story, the rest of which resides in a knowledge of context. The material consequences of the connoisseur's esteem, as we have been able to document them, are calamitous to the archaeological interest, because they have led to so many Cycladic figures being taken out of context that many roads to a secure knowledge of the prehistoric Cyclades, as once available in the archaeology of these figures, have been closed.

And it is impossible for the words and the ideas in which a modern esteem for Cycladic figures are commonly expressed not to be cast back to divert and to cloud our grasp of the realities of human life on the Cycladic islands, of which the secure evidence of figures and their contexts is so great a part.

It is evident how great are the intellectual consequences that follow and follow from the material consequences of esteem for Cycladic figures. Bereft of secure findspot, context, associated objects or date, they have only been made intellectual sense of by abstract frameworks of idealized canons and wished-for Masters in imagined studios. Yet intellectual consequences in turn create the material consequences. It is because the intellectual framework of Cycladic study so much depends on characteristics of the thing itself that each new figure that 'surfaces' without trace is regarded as of value, with just a passing aside or footnote to notice its lack of history, our lack of knowledge by which to reach understanding. As the material creates the intellectual, so does the intellectual create the material.

The drawings are by Arthur Shelley. We thank Cyprian Broodbank, Jack Davis, Wolf-Dieter Heilmeyer, Andrew Sherratt, Michael Vickers and Todd Whitelaw for their comments and criticisms. A first version of this paper was presented by David Gill at the Lawrence seminar, 'On museums and collecting', at the Faculty of Classics, University of Cambridge, May 1991.

Tables