Shamanism in Classical Scholarship: Where are We Now?

There can be no doubt that the idea of shamanism has had a powerful attraction for important and influential classical scholars in the last century. Yet this attention to shamanism was always limited to the small group of classicists that were interested in anthropology and the connections of the classical world with areas beyond the Mediterranean. In my contribution I intend to look again at the most interesting representatives of this interest – Diels, Meuli, Dodds, Burkert – in order to better trace the historiographical development but also note the problems that their proposed solutions raise and the answers that have been given so far. The end result should be a new determination of the status quaestionis today. As I look for the genealogies of the study of ancient shamanism, it might be useful to start with the moment when the shaman first became visible in Western Europe. What is the basis from which the classical scholars started to work? How did the concept of “shaman” find its way into classical scholarship? This problem has been treated by several scholars in the last two decades, but not without some confusion. Fortunately, two recent Groningen dissertations enable us to reach a better picture of the milieu in which Western Europeans became acquainted with the fascinating figure of the shaman.1 I will start by looking at a trail-blazing article of Carlo Ginzburg (§ 1), then analyse the classical scholars who have connected Greece with shamanism since the end of the 19th-century (§ 2), take a closer look at Aristeas of Proconnesus, one of the showpieces of the thesis of Greek shamanism (§ 3), and conclude with some considerations as to where we are now (§ 4).


In memoriam Walter Burkert
There can be no doubt that the idea of shamanism has had a powerful attraction for important and influential classical scholars in the last century. Yet this attention to shamanism was always limited to the small group of classicists that were interested in anthropology and the connections of the classical world with areas beyond the Mediterranean. In my contribution I intend to look again at the most interesting representatives of this interest -Diels, Meuli, Dodds, Burkert -in order to better trace the historiographical development but also note the problems that their proposed solutions raise and the answers that have been given so far. The end result should be a new determination of the status quaestionis today.
As I look for the genealogies of the study of ancient shamanism, it might be useful to start with the moment when the shaman first became visible in Western Europe. What is the basis from which the classical scholars started to work? How did the concept of "shaman" find its way into classical scholarship? This problem has been treated by several scholars in the last two decades, but not without some confusion. Fortunately, two recent Groningen dissertations enable us to reach a better picture of the milieu in which Western Europeans became acquainted with the fascinating figure of the shaman. 1 I will start by looking at a trail-blazing article of Carlo Ginzburg ( § 1), then analyse the classical scholars who have connected Greece with shamanism since the end of the 19th-century ( § 2), take a closer look at Aristeas of Proconnesus, one of the showpieces of the thesis of Greek shamanism ( § 3), and conclude with some considerations as to where we are now ( § 4).

The Introduction of the Term "Shaman" 2
In 1992 Carlo Ginzburg published an erudite article The Europeans Discover (or Rediscover) the Shamans, which has deservedly been translated into many languages and has found its definitive form in his recent book Threads and Traces. Ginzburg does not mention the place of first publication of his article, but it is not irrelevant to note that these pages first appeared in a kind of Gedenkschrift for the already mentioned Karl Meuli ( § 2), which was the result of a conference on Meuli in Basel, his hometown. In his contribution, Ginzburg argued that the first to mention the term "shaman" was the Dutch merchant Evert Ysbrants Ides (1657-1708/09), who registered the existence among the Siberian Tunguses of a "schaman or diabolical artist", 3 a word of debated etymology but certainly occurring only among Tunguse-speaking peoples. 4 The son of a Dutch immigrant in the Danish town of Glückstadt, in modern Schleswig-Holstein, Ides had founded a merchant house in Moscow, to which he regularly travelled starting in 1677. Here, in 1691, he met Czar Peter the Great (1672-1725), who, the following year, entrusted him with a missi on to the emperor Kangxi of China (1654-1722), with whom he had to initiate commercial contacts and to establish a more precise border between China and Russia after the Peace Treaty of Nerchinsk of 1689. 5 For his journey, which followed in a zig-zag route the northsouth course of the Russian rivers, Ides made use of a map of Siberia which had been made by the Dutch merchant and mayor of Amsterdam Nicolaes Witsen (1641-1717). Probably, Ides had received this map from Witsen's distant cousin and friend Andrej Winius (1641-1717), a member of the circle around the Czar. In 1632, Winius' father had moved to Russia to found a water-powered ironworks in Tula on the Tatar frontier where he married a Russian wife, but he had his sons educated in both Dutch and Russian. 6 Winius was an important source of information regarding Russia for Witsen and had acted as his interpreter during Witsen's visit to Russia in 1664-1665, but the latter carefully omitted the name of his cousin from all his books so that the latter's position at the Russian court would not be compromised. 7 After a trip of 18 months through Siberia and Mongolia Ides and his mission of more than 250 noblemen, advisors, merchants and soldiers reached Bejing in 1693.
Ides' main achievement was that every three years the Russians were allowed to do business in Bejing with a caravan of at most 200 members. However, he did not write his own account of the journey, but gave his papers to the already mentioned Nicolaes Witsen, who finally published the report in 1704 at his own expense. 8 It is not clear why this publication took so long. In 1697 Leibniz was already acquainted with Ides' expedition and mentioned Witsen in this context 9 -an acquaintance less surprising than it might seem at first glance: since 1694 Leibniz and Witsen had been corresponding with each other in French on a wide range of subjects, also on Witsen's Noord en Oost Tartarije (below). 10 In any case, Ides' report was a great commercial success, and within a few years it was translated into English, German and French and even, albeit somewhat later, Czech. 11 Strangely, though, Jacobus Scheltema (1767-1835), in his important study of the relations between Russia and the Netherlands, states that the Dutch edition did not appear before 1710, 12 although some copies have 1704 on the title page and the first translations appeared before that date. There seems to be something enigmatic about these early writings on Russia. However this may be, in the book Ides described the Tunguse shaman and provided the first illustration of a shaman in action. 13 Yet Ides was not the first to mention the term "shaman". As I discovered when studying Ginzburg's article, Ides had been pre-empted by the secretary of his expedition, Adam Brand, a merchant from Lübeck, but perhaps of Dutch origin, 14 who published a brief report of the expedition in 1697 in German. 15 This re port proved to be extremely popular in Western Europe, and Leibniz incorporated its contents in his Novissima Sinica. 16 17 The book was also used by Witsen for his edition of Ides' report, as the latter had not always supplied the exact dates in the course of the expedition. 18 Brand mentio ned that "where five or six Tungu ses live together...they keep a sha man, which means a kind of priest or magician". 19 I concluded, then, that in 1698 Europe ans could read the word shaman for the very first time.
However, at the very same time that I published my brief study of Greek shamanism, the English historian Ronald Hutton published an attractive book on shamanism and the Western imagination. 20 Given his interest in New Age beliefs and practices, it is not difficult to see why Hutton was interested in the subject, although it may have also helped that he is of Russian ancestry. In his preface he notes that the word "shaman" was "apparently first printed in the (Russian) memoirs" of one of the founders of the so-called Old Believers, Avvakum Petrov (ca. 1620-1682), but "seems to have reached Western and central European scholarship twenty years later in the works of Nicholas Witsen". 21 Regarding Avvakum, this is almost true, as Avvakum indeed uses a form of the verb "to shamanize" (shamanit), although not the noun "shaman". 22 Regarding the latter, Hutton is still somewhat doubtful, as he writes "seems to have reached".
However, all doubts have disappeared in Kocku von Stuckrad's study on shamanism and esotericism, which appeared shortly after Hutton's book and my own; in fact, having been a member of his Habilitation committee I alerted him to both studies, which he had not yet seen at the time. Von Stuckrad notes: "Jan Bremmer (…) bezeichnet Brand 1698 als die erste literarische Erwähnung des Begriffs (of the shaman); dieses Privileg kommt jedoch tatsächlich Witsen 1692 (1705) zu". 23 Evidently, Von Stuckrad took his cue from Hutton but, like Hutton, he overlooked an important aspect of Witsen's book.
Witsen published the first edition of his famous book Noord en Oost Tartarije in Amsterdam in 1692. 24 Its circulation must have been very low, as there are only four copies surviving, two in St. Petersburg, one in Utrecht and one in Amsterdam. 25 The reasons for this scarcity remain obscure but the fortune of his second edition was not that much better. Although the title page carries the year 1705, the first copies of the second edition appeared on the market only in 1747, thirty years after Witsen's death, once again for reasons that have not yet been clarified. 26 But whatever the reasons, for us it is much more important to note that there are significant differences. The second edition was wholly remade and expanded on the basis of additional and up-to-date information, such as the report of the expedition of Ides. Moreover, the plates in Witsen's second edition are often the same as those in Ides' book, and seem to derive from the same designer and printer who, unfortunately, have not yet been identified. 27 Von Stuckrad, who clearly did not make the effort to compare the two editions, thus missed the differences between them and wrongly credited Witsen with the first mention. 28 At the same time, we should note that the emergence of the term "shaman" has to be located in a specific network in Amsterdam around 1700 with Nicolaes Witsen as the spider in the web. It was his personal network and commercial interests that had promoted the production of Russian maps and reports of journeys in Russia. The discovery of the "shaman" was a fortunate by-product of this interest, but the figure was strange enough to become soon a focus of interest in the European Republic of Letters. 29

Rohde, Diels, Meuli, Dodds, Burkert 30
After Brand, Ides and Witsen, it would take some time before shamans entered the classical world. The first to compare Greek figures to shamans was probably Herder (1744-1803), who in a 1777 essay about the similarity of the older English and German poetry wrote that Arion, Orpheus and Amphion would have been "edle griechische Schamanen" when they lived. 31 This is still how the Romantic Movement pictures shamanism, but we come closer to modern views in that masterpiece Aglaophamus of the very learned but also very critical Christian August Lobeck (1781-1860), who in a note in his book argued that, if we wanted to, we could call Epimenides a priest, just like the Pythagorean Apollonius of Tyana and, as he notes in a surprising comparison, hodie sunt Schamani. Lobeck even adds that re vera they were just as much priests as Empedocles aut Abaris aut Pythagoras. 32 Unfortunately, Lobeck does not enlarge upon his comments or add references to literature regarding the shamans, but it is interesting for us to see that he already groups together several figures from the Archaic and earlier Classical period who we will also meet as a shamanic group in later writings.
Interestingly, there are two more unnoticed references to ancient shamanism in German classical scholarship that suggest that there was perhaps more attention to shamanism in connection with ancient Greece than we nowadays are inclined to suppose. In a mid-19th-century commentary on Herodotus, we find the shamans quoted in notes on Herodotus' description of divination practices of the Scythians. 33 The source of the commentary in this respect was the study of the ancient historian and geographer Karl Neumann (1823-1880), who had written his dissertation about Crimean Olbia. In a study of the Greeks in the land of the Scythians, the learned Neumann compares in detail some of Herodotus' information regarding Scythian divination and sacrifice, as well as the report that their women have two pupils in their eyes, with that of shamanistic practices of the Mongols, Buryats and other Siberian tribes. 34 Not surprisingly, we will meet his name regularly in the notes of Meuli's study on Scythian shamanism (below).
Although the nexus Greeks-Scythians-shamans has roots going back to the middle of the 19th-century, Wilhelm Radloff's 1884 Aus Siberien was probably the work most influential in animating interest in shamanism among leading classicists of the late 19th-century. Radloff (1837-1918) gave a fascinating description of a shamanic séance, a real "thick description", which still impresses by its attention to detail and liveliness. Yet Radloff was honest enough to stress that it is very difficult to give a precise definition of shamanism, as all tribes had variants of its beliefs and practices; moreover, because of the absence of written sources we have little idea of their history and authoritative traditions. Shamanism was of course less worthy than the three great religions of Christianity, Islam and Buddhism but it had its own value, and the shamans were not really less respectable than many Christian, i.e. Roman-Catholic, priests, according to Radloff. 35 Radloff's description of the shaman is rightly called an "unvergleichlich anschaulige Darstellung" by Erwin Rohde (1845-1898). In his masterpiece Psyche, Rohde does not really use shamanism as a comparative element in order to explain Greek phenomena, but he sees shamans as one manifestation of the masters of ecstasy who through their exhausting dances effect a "besonders energischer Glaube an Leben und Kraft der vom Leibe getrennten Seele des Menschen". In the end, it is this ecstasy, which is not exclusively shamanic, that Rohde sees at the basis of man's belief in the immortality of the soul. Rohde also adduced shamans to explain the lack of pain the maenads feel in Euripides' Bacchae, but again in this case the shamans are just one category among yogis, dervishes and the native North Americans. 36 Unlike Von Stuckrad, 37 we should therefore be reticent in attributing an all too great influence on later theories to him, even though Rohde had already collected virtually all the figures that Meuli, Dodds and Burkert would interpret as shamanic. 38 The second eminent classicist who adduced parallels from shamanism was Rohde's contemporary Hermann Diels (1848-1922).
In an 1897 article, he noted similarities between Anaximander's construction of the cosmos and that of shamanistic Mongolian tribes, but he also referred to the Kalevala, one of the shamanic texts which Meuli (below) was interested in. In his book Parmenides, which appeared in the very same year, he used Radloff's, as he called it, "klassische Schilderung", as the basis for his knowledge of shamanism although his interest went further than that. In rather Christianizing terminology ("Apostel", "Kirche", "Reformation", "Propheten", "Sündenfall" and "Sündenvergebung"), Diels grouped together a series of, as he called them, "Wundermänner", who indeed become visible in our evidence at the end of the Archaic period, around 500 BC. As with Lobeck, we find here, Abaris, Pythagoras and the Cretan diviner and purifier Epimenides, but also Parmenides himself, Empedocles, a certain Aithalides who had received the gift from Hermes that his soul could stay in the underworld and in the area above earth (Pherecydes B 8 DK), a Syracusan called Empedotimos of whom an ascent to heaven was told, the military commander Phormio about whom a visionary journey from Sparta to Croton, the hometown of Pythagoras, was related, and Aristeas, to whom we will return shortly ( § 3). In other words, in Diels we find the "usual suspects" of all those that are currently connected to shamanism.
However, it should be stressed that Diels did not accept shamanistic influence on ancient Greece. He saw the similarities more on a phenomenological than a genealogical level, 39 an attitude that was in line with that of Wilamowitz (1848-1931), the greatest classicist of that era, if not of all time, who was averse to any outside influence on his beloved Greeks. 40 That is perhaps why Diels had less influence than the two scholars who really put ancient shamanism on the map, Meuli and Dodds, to whom we turn now.
Meuli (1891-1968) was both a professor extra or dina rius at the Univer sity of Basel (and from 1942 an ordinari us) in Classics and Folklore as well as a teacher of Classics at the local Humanistisches Gymna sium. 41 His learned oeuvre stands at the crossroads of classics, folklore, ethnol ogy, psycho logy and the history of religion, and still impresses by its mastery of the sources and its careful, elegant style of writing. Unfortunately, we do not know how exactly Meuli came to shamanism. In my 2002 book I was still inclined to ascribe the main influence in this respect to Rohde, as Meuli had followed lectures in Munich in 1911-12, amongst others, with Rohde's biographer Otto Crusius (1857-1918), whom he highly respected. 42 Yet the absence in Rohde of any detailed reference to specific Greek shamans, makes me now realise that Meuli must have been especially inspired by Diels' Parmenides, whose study of Parmenides he calls a "meisterhafte Untersuchung", and whose characterization of Radloff's description as the "klassische Schilderung" he explicitly quotes. 43 Meuli's interest in shamanism becomes visible first in a passage from an article from 1924 on the bath of the Scythians, but it came to full fruition only in 1935 in his classic article Scythica, in which he concentrated on three aspects: 1. the Scythian shamans in Herodotus; 2. the transvestite Scythian seers; 3. the shaman and his poetry. 44 When we now read the article with critical eyes and do not let ourselves be swept away by its beautiful style and persuasive rhetoric, we can only conclude that in all three cases Meuli could only succeed in proving the existence of Scythian shamanism and the influence of that shamanism via several examples of sleights of hand.
In his discussion of Herodotus' description of the Scythians' funeral customs, Meuli focuses on the howling of the Scythians in their vapor bath with hemp. Quoting a description by Radloff of a shamanic purification of a yurt, Meuli interprets the howling as the singing of the shaman in order to guide the soul of the deceased to the beyond. Now Meuli was too honest a scholar not to ob ser ve that in Herodo tus' description all classic characteristics of shama nism are lacking: there is no mention of spirits, no mention of a drum, an indis pensable part of Siberian shamanism, 45 and, above all, there is no mention of a shaman! Meuli therefore suggested that the Scythians did not yet have professional shamans but knew an older stage of shamanism, family shamanism ("Familien-Schamanismus"), which could still be observed among modern day Siberian peoples, such as the Goldi, Votyak and Ostyak, or, as they are called today, the Nanai (or Nanay), the Urmurt and the Khanti (or Khanty). Yet among all these tribes hereditary shamanism is well attested. 46 When we now also take into account that Meuli himself notes that the ritual is described by Herodotus with great accuracy, 47 one cannot but conclude that he failed to substantiate his thesis at this point.
As regards his second point, Meuli focuses on the Enarees, of whom Herodotus (1.105, 4.67.2) relates that they were a Scythian group of seers who were the descendants of those Scythians that had plundered the temple of Aphrodite in Ascalon and were punished by the goddess with the "female disease", that is, made impotent. Consequently, as we hear from the author of a Hippocratic treatise (Aer. 22), they dressed in female clothes and performed female tasks, such as weaving, which looks like a Greek elaboration of the "female disease". 48 The Hippocratic author calls them Anarieis, which comes closer than Enarees to the undoubtedly Iranian origin of the name, which should not be explained, as Meuli does, from Iranian *anar, "not having a man", but from anarya:h, "not masculine". The variation Anarieis/ Enarees, with the typical Greek a>e adaptation of Iranian names, probably suggests that the Greeks learned of these seers along different routes. 49 Not surprisingly, Meuli compares them with reports of Siberian males who acted as shamans in female clothes, although not always without having sexual relations with women. Although Meuli himself notes that the closest parallels occur among the most eastern Siberian tribes, 50 this does not prevent him from declaring that "Die existenz eines echtskythischen Schamanismus ist damit bewiesen". Moreover, by combining the two Herodote an passages about the funeral and the Enarees, Meuli first promotes all the howling Scythians to being shamans and then also promotes the Enarees to a shamanic status, whereas Hero dotus describes them only as effeminate seers belonging to the elite. 51 However, despite his claim that apparently all Scythians were shamans, Meuli must have felt a bit uncertain about the lack of shamanism in what we know about the ancient Iranian religion. 52 This led him to deduce the journeys of the Mithraic initiates into the otherworld as closely related to those of the "primitiven Schamanen", a most unlikely suggestion. 53 I conclu de, therefore, that at this point, also, Meuli has not proven the existence of Scythian shamanism. 54 His final point concerns shamanism and Greek poetry. After an interesting account of shamanistic poetry, Meuli first discusses Aristeas and Abaris, figures already identified as shamanoid by Lobeck and Diels, as we saw. 55 I will come back to Aristeas shortly ( § 3), but note that Meuli promoted both of them to the status of "skythischer Wundermann", although there is no evidence at all in our sources that they were Scythians. 56 The only other figure Meuli considers as shamanic is Zalmoxis, a Thracian whom the Greeks associated with Pythagoras in a manner that has not yet been satisfactorily explained and will not occupy us here. 57 Having looked at these figures, Meuli concluded: "Die Existenz einer skythischen Schamanendichtung, die bei so entwickeltem Schamanentum ohnehin anzunehmen war, darf nun als erwiesen gelten". This is of course a rhetorical trick, as Meuli had proved neither the existence of Scythian shamans nor the existence of Scythian shamanic poetry. But he needed this conclusion in order to reach the point he wanted to make, that is, that the journeys of Odysseus and the Argonauts in Greek epic went back to shamanistic poetry. Needless to say, a sober look at the evidence does not provide any proof of these suggestions despite all Meuli's erudition.
Von Stuckrad quotes this conclusion, too, and remarks: "Diese Darstellung ist der germanischen Rhetorik der 'Herrenrasse' ebenso verpflichtet wie den Konstruktionen des edlen griechischen 'Geschlechtes', die sich dem 19. Jahrhundert verdanken". 58 Nothing could be further from the truth. Meuli had nothing to do with the contemporary Nazi ideology, and his whole oeuvre testifies to a view of Greece that does not stress its superiority but its indebtedness to Central Asiatic traditions. In 1940, in support of the Finnish war effort against the Russians, Meuli returned to the problem of shamanism in a brief piece on the Kalevala, in which he postulated shamanic poetry as its core, and in a lecture given in 1950, but which was only published in 1975, he postulated Orpheus as an "Urbild eines Schamanen". He remained interested in shamanism until the end of his life, 59 but his work became really influential only through the use made of it by Dodds.
Unlike Meuli, Eric Robertson Dodds, Regius Professor of Greek at Oxford, (1893-1979), 60 invested most of his scholar ly time in books, three of which -his com mentary on Euripides' Bacchae (1944), The Greeks and the Irrational (1951), and Pagan and Christi an in an Age of Anxiety (1965) -are still being reprin ted and translated, thus making him the most in fluential English classicist of the 20 th -century in internatio nal terms. Yet it is especially his book about the Greeks and the irrational that has been the most influential. This is also the book that popularized the notion of the shaman well into the twenty-first century. However, Dodds' use of the idea of shamanic influence was completely different from that of Meuli, as he wanted to explain the rise of the notion of the immortal soul in Greece, which he saw as something new and alien to Greek culture. That is why he latched on to the idea of shamanistic influence. Dodds thus accepted Diels' and Meuli's arguments without any criticism, and even expanded Meuli's collection of shamanic figures by incorporating Pythagoras, Empedocles and, without knowing Meuli's posthumously published study, Orpheus, whom he interpreted as "a mythical shaman or prototype of shamans". 61 It is rather odd that Dodds explains the growth of a certain Puritanism, which he perceives in Archaic and early Classical Greece, as the result of the impact of shamanistic beliefs. The little that we know about Scythia and Thrace gives no information at all about shamanistic beliefs. It is clear that, basically, Dodds was at a loss as how to elucidate the rise of the independent soul, a problem that has indeed not yet been satisfactorily explained. 62 Apparently, there are too few data to lead us towards a specific explanatory path and the available traditions can be juggled into several directions without our evidence giving us sufficient guidance.
It is noteworthy that Walter Burkert (1931-2015) realised this problem to a certain extent. He had got to know Meuli through Reinhold Merkelbach , who was well to do and had been able to restore contacts with Switzerland fairly early after the war. Merkelbach, then professor of Greek in Erlangen where Burkert was Assistent at the time, was an admirer of Meuli, and one of the driving forces behind the publication of the latter's Gesammelte Schriften. Burkert shares this admiration, as also appears from his famous book on sacrifice, Homo Necans (1972). It is therefore not surprising that Burkert was inspired by Meuli's study of shamanism, although he of course also admired Dodds and Diels, whose Kleine Schriften he edited. Following Diels, Dodds and Meuli in his Habilitationsschrift (1962) about Pythagoras, Burkert not only put Pythagoras and the now "usual suspects", such as Aristeas and Abaris, in a shamanistic framework but he also discussed the problem of Greek shamanism in a separate article, which has received less attention than it deserves. 63 In this article Burkert starts by wondering whether the adduced shamanic motifs, such as journeys to the Beyond, are simply "legendäre Wandermotive" or practiced rituals, either as survivals from time immemorial or revivals due to foreign influence. He does not answer the question but directs his attention to a word that he considers to have a noteworthy "Affinität zum Schamanenbereich", namely γόης, which he translates with the Scottish classicist John Burnet (1863-1928) as "Medizinmann". He then continues by arguing that the γόης was the closest equivalent to the Siberian shaman. But is that really true?
Burkert starts with the reference in the archaic epic Phoronis (fr. 2 Bernabé) to the Idaean Dactyls, smiths who were also γόητες, "sorcerers", and who engaged in "incantations, initiations and mysteries". Apparently, the combination with incantations is very important, as a number of passages in Greek literature combine these with sorcerers or sorcery. 64 Burkert sees the bridge between these two in Musikmagie and subsequently extrapolates from the calling up of Persephone in Eleusis with a gong and the fact that necromancers are connected with "sorcering" a close connection between the "sorcerer" and the cult of the dead. He also notes the report by Herodotus that each year the Scythian Neuroi morphed into wolves for a few days and naturally concludes that these people must be "sorcerers". Yet the fact that Herodotus locates these Neuroi next to the cannibalistic Androphagi suggests a certain fictional content in this report. 65 However, Burkert concludes from his examples that there were certain persons in olden times who were the centre of ritual and cult: "Medizinmänner, Schamanen, Zauberpriester". 66 Burkert proceeds with the etymology of γόης and concludes from the use of the corresponding verb γοάω and related adjectives that its basic meaning was "to lament". Consequently, he argues, the γόης was "the wailer", the performer of the funeral lament. Unfortunately, one must object: there is not a single passage in Greek literature where the term has the postulated meaning. Moreover, Burkert has also to resort to some hermeneutical juggling as he notes the absence of an ecstatic Jenseitsreise in the case of the γόης, despite the fact that such a journey to the Beyond "ein besonderes Charakteristicum des Schamanen ist", but suggests that the "oft hervorgehobene Kunst der Verwandlung in gewissem Masse äquivalent ist" -which is not immediately persuasive. One can therefore only agree with him that "Ein direkter Beleg für entwickelte Schamanenpraxis ist allerdings in unserer Überlieferung nicht zu finden", 67 and neither is an indirect one, I may add.
Moreover, it is not true that the negative meaning of γόης is a gradual development. From the very beginning, we can note a combination of the γόης with magic and incantations. It seems that the term, which etymologically means "the shouter", 68 developed in two directions. On the one hand, among women it became used for wailing and lamenting, whereas among men it seems to have been used for the loud performing of incantations. At least that is the most natural explanation, given its frequent combination with incantations (above) from the very beginning. In any case, the term always has a negative meaning and is never a self-designation. To postulate it as "zum Schimpfwort abgesunken" lacks any basis in our evidence. 69 I can only conclude that this attempt at introducing the shamanic model as a major hermeneutical tool for Greek religion has not been successful.
Burkert returned to shamanism in 1972 in the English version of his Weisheit und Wissenschaft. In the preface he notes that "more thorough acquaintance with ancient religion has pushed the concept of "shamanism" further into the background", and in his classic handbook of Greek religion, originally published in 1977, shamanism makes only a fleeting appearance. 70 Yet in his 1979 study on the Master of Animals he took up the theme again, but introduced a new twist. Now he explained Heracles' hunting of the herd of Geryon from a primeval shamanistic motif that can still be witnessed among Arctic and Siberian peoples in the recent past, and with roots going back to "the darkness of prehistory". 71 No more Scythians here! Shamanism has now become one of the strata of man's civilization of which certain strands survived into the historical period. It is not surprising, then, that in the "Addendum 2003" to the reprint of his 1962 article Burkert rejects my plea for "a more detailed definition of the shamanistic complex", and pleads for "shamanism" (his inverted commas!) as a way to explain Jenseitsreisen in connection with certain ritual practices. 72

Aristeas of Proconnesus
In his Ecstasies Carlo Ginzburg basically follows Meuli and looks for a historical connection between Greece and the Scythians to explain the shamanistic motifs in early Greek culture. He does not survey all figures traditionally connected to Greek shamanism, but just mentions Aristeas of Proconnesus. As he is also the only shamanic figure that has received special attention in the last decade, it may be worthwhile to have a second look at him in order to get the problem of Greek shamanism into better focus. 73 Let us not rehearse here the whole of his story but just the beginning as related by Herodotus: "Aristeas also, the son of Kaystrobios, a native of Proconnesus, says in the course of his poem that, possessed by Apollo, he reached the Issedonians. Above them dwelt the Arimaspi, men with one eye; still further, the gold-guarding griffins, and beyond these, the Hyperboreans, whose country extended to the sea. Except the Hyperboreans, all these nations, beginning with the Arimaspi, continually encroached on their neighbours. Hence it came to pass that the Arimaspi gradually drove the Issedonians from their country, while the Issedonians dispossessed the Scyths; and the Scyths, pressing upon the Cimmerians, who dwelt on the shores of the southern sea, forced them to leave their land" (4.13, tr. Bolton).
I have limited my quotation for the moment to these lines, as they are sufficient to establish certain data about Aristeas. The name of his father Kaystrobios means "gift of Kaystros", and the first part, Kaystro-, often occurs in Ionian names. 74 This firmly establishes his father as an Ionian who, presumably, had emigrated to Proconnesus, an island in the Sea of Marmara. His son Aristeas was the author of a poem. As we have a number of lines of that poem, the Arimaspea, the safest way of dating Aristeas is through an analysis of the words of his poem and their chronological occurrence in Greek poetry. Such an analysis was first carried out by Bolton in an excellent book on Aristeas and his analysis has subsequently been refined by Ivantchik. The latter reaches the conclusion that the language of the poem dates from about 500 BC, which comes close to Jacoby's placement of the poem in the second half of the sixth century BC. 75 The date is confirmed by the iconography of Greek vases where the battle between the Arimaspi and the griffins, which is described later by Herodotus, starts to appear around 515 BC; 76  The relatively late date makes it much more likely that Aristeas heard stories from the Scythian areas than the traditional seventh-century date would have allowed, as it is hard to think of a Greek traveling in Scythia in the seventh century. 77 Without a "Rough Guide", how would he have found his way in a land without landmarks and with inhabitants whose languages he did not speak? It is one thing to hear of stories about gold-guarding ants, but a rather different matter to become trained as a shaman and to be able to go into trance with a concept of the soul foreign to one's habitus. 78 Given that Herodotus certainly wrote about a century after Aristeas we should also wonder about the historical value of the traditions he recorded in Proconnesus, as these look very much like later embellishments. This is even truer for what he heard in Metapontum. Here, as Herodotus was told, Aristeas re-appears 240 years after his disappearance in Proconnesus and told the Metapontines that he followed Apollo in the shape of a raven. We need not go as far as Bolton and speculate that Pythagoras himself was especially interested in the Arimaspea, 79 but the re-appearance after death, the closeness to Apollo (the main god of Metapontum) and the metamorphosis into a raven, Apollo's bird, clearly all point to a Pythagorean background with Metapontum as centre. It is not surprising, then, that in the catalogue of the Pythagoreans reported by Iamblichus, which goes back to the fourth-century BC Aristoxenus, we find an Aristeas among the Pythagoreans of Metapontum. 80 In the end there is no early shamanistic detail left of Aristeas' legends that is credibly derived from the Scythians. 81

Where are we now?
It is perhaps not surprising that after the expansive views of Greek shamanism a reaction set in, first by myself and directed especially against the expositions of Meuli and Dodds, then by Fritz Graf against the shamanistic interpretation of Orpheus and, last but not least, by the late Pierre Hadot (1922-2010) and the Russian scholar Leonid Zhmud against Burkert's shamanising interpretation of Pythagoras. 82 Altogether these investigations have gradually dismantled the edifice built by Meuli, Dodds and Burkert and shown that none of the proposed derivations from the Scythians holds up or even that the Scythians themselves had a shamanistic religion. At the same time we could not but note that the shamanistic concept served very different purposes: for Meuli it explained the origins of epic poetry, for Dodds the origin of the immortal soul and the rise of charismatics such as Empedocles, and for Burkert the legends surrounding Pythagoras. In other words, shamanism was, so to speak, a joker that could be put on the table to explain developments for which scholars were unable to produce an internal Greek explanation. 83 Now where classical critics focused on the "diffusionist" approach, Von Stuckrad has approached the problem from a different angle. 84 He rightly argues that when scholars adduce modern descriptions of shamanism to explain ancient phenomena, they inevitably presuppose the unchanging character of shamanistic cultures, whereas everything we know suggests that these did not stand outside the flow of history. This is certainly true, although we also should observe that at the time of the supposed shamanistic influence on Greece, that is around 500 BC, Central Asia had not yet experienced the shock waves of Buddhism, Islam or Christianity.
On the other hand, it seems less helpful to speak, as von Stuckrad proposes, of a schamanistische Matrix, which consists of "Transformation, Jenseitsreise, Initiation, Heilung, Kommunikation mit Toten, Trennung von Körper und Seele sowie aussergewöhnliche Bewusstseinszustände, die in der Regel durch Musik induziert werden". 85 Von Stuckrad clearly fails to realise that he has arrived at his Matrix from the very modern descriptions he first considered unusable to interpret ancient phenomena. 86 Moreover, he pleads to use only emic concepts, which is never helpful, 87 and succumbs to the magic of the γόης by considering that figure a helpful tool for a better understanding of the "shamanistic" figures without noting that it is a term of abuse, not a self-designation, and that there never was a recognizable figure in Greece that matched his Matrix.
A rather different direction was taken by the great French scholar Louis Gernet (1882-1962), a member of the Durkheim school, in a generally neglected article of 1945. He looked at the same "shamanoid" figures as Diels and Meuli had investigated, but tried to illuminate the various traditions from the pre-history of the Greeks, even extending his explanatory framework to the Indo-Europeans. Thus Gernet explained the connection of Pythagoras with the divine from Frazer's magical king of the Golden Bough. Gernet's contribution is without notes, but the brief remark that the notion of the soul that has been picked up by Platonism was once associated with "quelque chose comme une discipline de shaman" leaves little doubt that Gernet also knew Meuli's article. However, Gernet realized that he spoke more of "antécédents plûtot qu'un passage". 88 One need not follow Gernet in his Frazerian approach, but it is indeed a fact that the mythical figure of Orpheus and the activities of the Cretan Epimenides (firmly dated to about 600 BC) suggest that the roots of these figures go back into times that are no longer accessible for historical research.
Unlike Gernet, his pupil Jean-Pierre Vernant  was the first to attack this problem in a 1959 article, entitled "From Myth to Reason", in which he tried to explain the "shamanoid" figures without taking recourse to the shamanic paradigm, although even he could not wholly escape the fascination of the shaman. Vernant notes the high social status of the first philosophers, coming from priestly families, who now put their priestly secrets in the open air. More importantly, he connects the transition from the Archaic Age to the classical era with the combined birth of the polis, the growth of politics, the rise of mathematics, and the emergence of money. 89 Although seemingly without knowledge of Vernant's article, Burkert refined this line of thought in his γόης study. He persuasively argues that the polis no longer had room for the exceptional individual, and, as I have argued, it is this reduction of possibilities for the nobility to stand out in this life that promoted a belief in reincarnation, which would still guarantee them a special status, even if perhaps only in the life hereafter. 90 Yet, in contrast to Gernet's observation, these arguments may explain their disappearance, but they do not illuminate the antecedents of the "shamanoid" figures.
So where are we now after the dismantling of the shamanic paradigm? We have seen how the best Hellenists of the last century have wrestled with the problem of the appearance in our sources of people who were reputed to be able to go into a trance and to fly, to practice a special life style or to possess extraordinary knowledge. Their interest in this theme could make them members of a Faszinationsgemeinschaft as defined by Martin Mulsow: 91 whatever their solution, all these scholars were clearly fascinated by the charismatic outsiders, who are so different from the rational Greeks as we like to see them, and their fascination led them to sometimes uncritical acceptance of the shamanistic theses. Yet their studies also have greatly elucidated the various traditions of the "shamanic" figures and their mutual relationships". 92 This conclusion does not preclude the possibility of cultural and religious transfers between Central/South East Asia and Greek culture, not even an influence from shamanistic cultures. Aristeas' poem told of journeys to fabu lous peoples and of gold-guarding griffins fighting with the one-eyed Arimaspi, a passage that surely is a dou ble of Herodotus' report about gold-guar ding ants in the Bactrian desert. The story is well attested in ancient Indian sources and probably derives from Dardistan where the burrowing of marmots in the gold-bearing soil was regularly exploited. 93 In his recent book on Indo-European myth and poetry, Martin West has suggested that the Greeks and Indians derived their ideas of reincarnation from a common source somewhere in the Persian Empire. Yet if we take into account that the Buddha has been down-dated in recent decades, the chance is not imaginary that the Indians were, directly or indirectly, influenced by the Greeks in this respect. 94 West also compares the close parallels between the role of the raven in the myths of Germanic Odin, Celtic Lug and Greek Apollo and ascribes them to the influence of Finno-Ugric peoples in different directions. Perhaps this is possible, but one must also observe that Odin and Lug have two ravens, Apollo just one. In fact, Apollo really appears too late in Greek religion to be persuasively compared with Odin and Lug, the more so as he probably derives from pre-Indo-European Western Anatolia. 95 The existence of a kind of world pillar, on the other hand, looks specific enough to be derived from Central Asiatic cosmology, but we do not know when that happened. 96 In the end, though, it is easier to accept that the Greeks derived some mythological details from Central Asia than that they imported complex ritual practices that presuppose a new concept of the soul.
In reaction to the shamanistic approach, Gernet, as we saw, attempted to situate the "shamanoid" figures in the heritage of the Greeks from their religious prehistory, even if with many permutations in the course of time. On the other hand, more recent scholarship has not looked into the Greek hoary past but has concentrated on internal developments at the time of the transition of the Archaic to the classical age. It seems to me that future research should try to combine all three approaches. There can be little doubt that the Greek "shamans" did not appear from nowhere but made use of practices and ideas that had "une très longue histoire", to quote Gernet one last time. 97 At the same time we need not exclude the possibility of religious transfers from the Thracians or Scythians, but we should be more rigorous in our explanations than Meuli and his followers have been. Finally, we should try to trace the political, cultural, economic and religious developments that created the world in which these figures could operate but from which they also disappeared. This is not an easy task, and we must somewhat sadly conclude that, despite all the efforts of the best classical scholars of the last century, we can still see these Greek "shamans" only through a glass darkly. 98