Josipa Lulić
There is an unusual group of reliefs in the Roman province of Dalmatia. They are funerary reliefs (those that have a known archaeological context, some are accidental finds), and they depict Mercury. However, they break away from his classical iconography. Usually, Mercury is depicted holding a caduceus (Greek kerykeion, a herald staff that branches into two intertwined twigs; sometimes understood as two snakes) and a money bag, dressed only in a cloak, and wearing winged hat and sandals. While Mercury is in possession of most of his usual attributes, there are two rods in his hands: the familiar caduceus in one, and a second rod in the other hand, the one usually reserved for the marsupium (the money bag). That is not unique in the Roman Empire, but it is quite exceptional: there are only a handful of comparative examples, and the Dalmatian group is the only one that is geographically and chronologically coherent. I propose to investigate possible iconographic solutions and influences for such a depiction, but there is a second aspect, which is just as important – the anthropological meaning of this imagery. On the one hand we have the religious and philosophical underpinnings of this iconography, on the other the role that such an image may have played in the religious community of a certain time and place. How did the patrons and artists that produced the sculptures understand the meaning of this visual sign? The History of Art focuses on the object and on possible similar artistic themes from other times and places; anthropology focuses on the human community and on the possible role of the object in a system of cognition and communication. Thus, this chapter will explore this iconography from two different angles. First, it will locate the possible iconographic meaning of the motif of Mercury holding the second rod; second, it will, through the use of cognitive theory, try to answer the question of what may have been understood from this motif in Roman Dalmatia.
There are more than twenty sculpted depictions of Mercury in Roman Dalmatia, which amounts to approximately 8 per cent of the total number of religious sculptures in the province (Lulić, 2015, p. 157). The number of epigraphic monuments is somewhat smaller; there are only a dozen inscriptions mentioning Mercury in Dalmatia. Mercury is not the most common deity in the corpus of religious images, but he appears much more frequently than, for example, Jupiter or Juno, even if we take into account only sculpture in stone (marble and local limestone): small bronze figurines are frequent in Dalmatia, as they are in all of the Roman Empire – especially in large ports such as Salona (Barr-Sharrar, 1990, pp. 210–215). In that corpus, we can isolate a group of six depictions of Mercury with characteristics in common. They are all reliefs in local limestone, in which Mercury is depicted as a young man, draped only in a cape, which is thrown over the deity’s left shoulder. In his left hand he holds a caduceus. He does not carry the money bag we might expect, but he does wear a winged hat on his head. His body is positioned facing the front, sometimes leaning slightly on his right leg. The most striking characteristic of the depictions in this group is that Mercury holds a second rod in his right hand, which is, unlike the caduceus, plain and straight. All of the sculptures were produced locally, and in all cases where we know the context, the item had a funerary function.
The earliest example is a sarcophagus fragment from the Narona cemetery, kept in the Archaeological Museum in Zadar, dated to the early 2nd century CE (Cambi, 1980, p. 137, tab.17; Hirschfeld and Schneider, 1885, pp. 50–51; Patsch, 1899, p. 505, fig. 1; Patsch, 1900, p. 93, no. 1, fig 65; Sanader, 1986, p. 116, no. 140, fig. 140; Sanader, 1994, p. 90, no. 9). On this fragment we can see two arcades with undecorated arches stemming from plain-shaft pilasters, with capitals decorated with schematized acanthus leaves. On the far right-hand end of the fragment, above the arch, there is a vegetable motif, which points to it being the end of the arcade, and probably the far right edge of the front side of the sarcophagus. Under the arches there are two male figures on pedestals. The first is a naked, beardless male figure of strong musculature, with a lion skin over his left shoulder and a leash in his right hand; here we can easily recognize Hercules capturing Cerberus (Cerberus has not been preserved). In the second arcade we see a naked male figure in a travel cloak, with a caduceus and a winged hat (petasos), easily identifiable as Mercury. Both figures are in low relief, standing on pedestals (suggesting that they represent statues depicted in relief). If we were to reconstruct the sarcophagus with five arcades equally wide on the front side, the width of the sarcophagus would come to approximately 2.5 m, which corresponds to characteristics of a typical columnar sarcophagus (which usually has an uneven number of arches, usually five), as well as to the average size of Dalmatian sarcophagi (Koch, 1933, pp. 27–32; Lawrence, 1932, p. 150; Cambi, 2010).
The sarcophagus was most likely produced in the local workshops of Narona. Although sarcophagi from Salona (both imported and locally produced) are well known, we only have a few examples from Narona, since the swampy terrain under which the necropolis lies prevents archaeological research (Cambi, 1975, 1980, 1988, 2010). There is no preserved inscription on the fragment, but the patron of the sarcophagus must have been a wealthy person. The only known source for determining the price of sarcophagi is an inscription on a sarcophagus from Salona (Cambi, 2010, p. 46): in the late third century a person would have to pay fifteen gold coins (solidus) for a medium size sarcophagus without decoration, an amount of money that would provide basic necessities for five years (Russell, 2010, p. 122; Jongman, 2007). The prices obviously varied depending on whether the sarcophagus was locally produced or imported, and how rich the decoration was. Most sarcophagi would have been ordered from the shelf, already half-finished, with a typical array of motifs that were in fashion at the time (such as the seasons, or some abstract decorations with a rectangular frame containing an inscription, tabula ansata). The Narona sarcophagus does not look like any of these. We can also assume that a sarcophagus with atypical motifs, which required a special order, was more expensive than one that could be sold directly from the warehouse. In this case, the choice of a local workshop might have been made due to the need to supervise the production and arrange the iconography of the scene directly (Toynbee, 1996, p. 274; Russell, 2010, p. 122). This implies that the patron(s) had a clear idea of what they wanted for their final resting place, so the imagery was not chosen at random, but probably expressed some of their beliefs about life after death. I do not suggest that the iconography of the sarcophagi in every instance needs to be studied in a religious key, as a portal to the ideas about the afterlife, – sarcophagi were sculptured to embody moral virtues, as well as to serve as visual therapy for the survivors among other things (Platt, 2011, pp. 340–341) – but the explicit choice of religious iconography instead of more typical motifs suggests that this may have been the case.
Other than on the Narona sarcophagus, we find Hermes/Mercury with petasos and caduceus on a fragment of the short side of a sarcophagus from Salona, but the damage to his right hand prevents us from firmly concluding if he also had a short rod, though this is implied by the position of the hand – it is an exact iconographical match to the Mercury from Narona (Cambi, 2010, p. 119, no. 111, tab. LXIV, 2). Eros is depicted on the same fragment, creating a symmetrical composition. In Split museum there is another depiction of Mercury from Salona, with the same attributes, and it is even more similar in his body position to the Narona fragment (Bulić, 1885, p. 41, no. 120; Cambi, 2010, p. 115, no. 91, tab. LIII, 2; Cambi, 1960, CIL III 13943 (9291)). This is the front of a limestone sarcophagus, with a tabula ansata and small children, putti or eroti, shown as personifications of seasons on the left-hand side. This arrangement can often be found in the corpus of decorated sarcophagi, and is therefore something that could have been picked out from the shelf. That makes it even more interesting when at the bottom of the fragment, next to the feet of the figure, we find a small relief depicting Mercury with all of the attributes already mentioned in the above examples (and an extra, a rooster by his feet). While personifications of the seasons and putti are common-place in sarcophagi produced in the Roman Empire (Lawrence, 1958), the interjection of Mercury into a scene which has no iconographic connection to him suggests that the patron has intervened in adding Mercury to a generic theme. Another Mercury in a funerary context can be found on a small pillar in a funerary precinct near Salona, with Eros on the corresponding pillar (Abramić, 1932, pp. 62–63, tab.VI, fig. 4). Other examples of Mercury with a second rod from Roman Dalmatia only have brief notes about a general area where villagers had found the sculptures and brought them to the nearest interested person – we cannot determine their context, since there are still large potential archaeological sites awaiting archaeological surveys (Patsch, 1914, p. 194, fig. 82; Patsch, 1894, p. 54, fig. 3; Imamović, 1977, p. 390, no. 140, fig. 140). The Narona fragment is the earliest example, and it was the most deliberate one; the only one where Mercury Psychopomp is shown with other figures (such as Hercules in his task of taking Cerberus from Hades) which could indicate the theme of a journey through the underworld.
This iconographic particularity was recognized in the scholarship relatively early (Patsch, 1900, p. 93), but it has not yet been interpreted. The only scholarly mention of the strange iconography of the Dalmatian group comes in the form of the attribution of the type to Hermes Psychopomp, without closer analysis (Imamović, 1977 p. 390; Patsch, 1894 p. 55). This lack of interest might be because it does not fit in with the main paradigms for the study of religion in Dalmatia, those of Romanization and resistance (Lulić, 2016). Both paradigms assume fixed ethnic identities, clearly reflected in the religious sculpture, and they are trying to understand their relationship in terms of power negotiations. The Romanization paradigm thus assumes that the Roman religion took over the native one, while the resistance paradigm looks for clues in the iconography in order to prove that the native ethnic identity remained intact during the Roman rule. The image of Mercury does not neatly fit to either. In this case we need a different paradigm, that of the formation of the provincial habitus (in Bourdieu’s terms) beyond the discussion of Romans and Natives.
There are other depictions of Mercury holding a second rod from Greek and Roman antiquity, but they are relatively rare: there are more examples from second and third century AD Roman Dalmatia than all other known examples from Greco-Roman antiquity combined.
The earliest example is from classical Greece is a lekythos from Athens. On it, Hermes holds the kerykeion in his left hand and in his right a simple rod. In front of him two winged souls fly out of the funerary urn, while a third one is just making its way out of it (Schadow, 1897, pp. 16–17). The second example is from a funerary altar in Milan, Italy, where we see Hermes/Mercury holding a caduceus in his left hand, and a simple rod in the right hand (Dütschke, 1882, p. 398, no 970). On the opposite side is Hades, and Charon is depicted on the rear side. There are depictions of Mercury with a second rod in Rome as well: the one that is the most similar to the Dalmatia examples is a Mercury on a sarcophagus lid in the Capitoline museums (Schadow, 1897, p. 25). The lid is dated to the midsecond century and also features depictions of Hades and Persephone on the throne, as well as the three Parcae. There is a similar depiction of the deity on a statue of Hecate from Salinae (modern Ocna Mureș) in Dacia (Petersen, 1881, p. 67; Gramatopol, 1982, p. 132, Taf. III/15; Gramatopol, 2000, p. 256). On one of Hecate’s three sides there are numerous figures in multiple registers, with Helios in the centre of her chest. In the first register we see Mercury depicted almost identically to the examples from Roman Dalmatia, with animals on his left – a rooster, a horse and a turtle – and a woman with a veil and a dog on his right. The most recent example is from the Vibia catacomb in Rome (Ferrua, 1971, passim; Casagrande-Kim, 2012, p.165–170; Jastrzębowska, 1979, pp. 38–40. For a detailed exploration of this tomb, see Gabriela Ingle’s contribution to this volume). There, among the images of Vibia’s journey to the afterlife, is a painted scene of two chthonic deities on the throne, identified as Dis Pater and Aeracura, in the judgement scene:1 in front of the throne is Mercury holding the second rod.
As for the literary evidence, Hermes was already seen as a chthonic deity in the Greek world, as important a part of underworld imagery as Hades and Persephone (Waele, 1927, p. 31; Pausanias 1, 38, 7). The Homeric Hymn to Hermes (Homeric Hymns, 13) mentions his role as the bearer of dreams and the guardian of the night, the Orphic Hymn 57 to Chthonian Hermes invokes the power of his wand to wake the sleeping or the dead, and his role between the worlds is often highlighted in Homer (for example at Homer, Iliad, 24.333; Homer, Odyssey, 5.28; 24.1). He is also portrayed as a necromancer on Greek vases, using magic to invoke the Earth goddess (Waele, 1927, p. 56). Hermes was always portrayed carrying a magic rod that he used to navigate the worlds: it could be used to wake someone up, or to put them to sleep (Waele, 1927, p. 34). In Homeric poetry, the rod in question was consistently referred to as to the rhabdos: the same word that was used for Circe’s magic wand (Homer, Odyssey, 10.238), as well as for the one used by Athena to transform Odysseus’ appearance (Homer, Odyssey, 16.172). This word is also used by Pindar to describe Hades’ rod with which he leads the deceased (Pindar, Olympian Ode, 9.33), and by Herodotus when he describes the willow wands used for divination (Herodotus, 4.67). The term kerykeion was used for the first time in relation to Hermes in the writings of Herodotus (Hdt. 9.100) and Thucydides (Thuc. 1.53; Waele, 1927, p. 35).
In Greek religious thought, Hermes represents the deity of borders and magic; he is a mediator between the worlds (Benoît, 1959, p. 147; Gulizio, 2000, p. 113). Etruscan Hermes-Turms-Mercury seems to have fulfilled a similar role, his incarnation as a psychopomp connected to Orphic beliefs (Combet Farnoux, 1980 p. 366), probably through the influence of the colonies in Magna Graecia (Cumont, 2005, p. 5). Roman Mercury is primarily the god of commerce (Combet Farnoux, 1980, p. 336), although literary sources also refer to Mercury Psychopompus (Benoît, 1959, p. 164). His role as the guide of souls is often used in the context of discussions of death as a transition, of changing states (Combet Farnoux, 1980, p. 356). He is depicted in this light in the Aeneid, where Virgil mentions the power of his rod to wake up shadows and guide souls (Virgil, Aeneid, 4.242), which has sometime been interpreted as influenced by Orphic-Pythagorean beliefs in reincarnation and the transmigration of souls (see for example Combet Farnoux, 1980, p. 373, Waele, 1927, p. 57). In the written evidence as well as visual examples, Orphic ideas about the journey through the Underworld remain the recurring connection, and therefore require a closer look.
The word Orphism, as it is true for many other categories (some would say even the Roman Empire itself; Barrett, 1997, p. 53), is a recent creation, which helps us to describe and understand certain phenomena from the ancient world. But unlike the Roman Empire, which is still quite a useful construct, many researchers would argue that Orphism is not. In its linguistic formation, the –ism supposes a clear religious system with rules and rituals, but that is not the image we find in the sources. In the place of the religious system is a polythetic group, Orphic literature, or a climate of opinion (Alderink, 1981; Edmonds, 2004). Even more so, Edmonds claims that ‘Orphic’ was an epithet applied to any extraordinary religious tradition (Edmonds, 2004, p. 103).
When investigating possible connections between the Dalmatian material and Orphism, Alderink’s definition of Orphism as a climate of opinion may be the most useful, without forgetting Edmonds’ objections to the connection of any belief in the afterlife with Orphic tradition (Edmonds, 2013, p. 82). As an open system, or structure of thought – two other definitions Alderink has given for Orphism – Orphism included many phenomena, not all of which were part of what defined the Orphic way of thinking. Orphic ideas about the journey of the immortal soul through the underworld (as expressed in the Orphic gold tablets) were part of that climate (Bernabé Pajares, 2008; Bremmer, 2002). Aldernik’s definition is inclusive of the fluidity of ideas and concepts between Orphism and other religious groups and concepts, both directed to and coming from the Orphic climate of opinion (Alderink, 1981, p. 23). Orphic ideas found their way through different philosophical and religious systems from the Hellenistic period onwards, reaching their peak in the Roman Empire during the time the Dalmatian material was created, from the second century CE (Herrero de Jauregui, 2010, p. 31).
If we go back to the examples found in other times and places, they are closely connected to the chthonic aspects of Mercury. Most of them were also found in funerary contexts: the Jena vase was an urn, the Milano relief comes from a grave and the two Roman ones are from a sarcophagus and from a catacomb. They also refer in their iconography to the Underworld, with representations of Hades and Persephone. The catacomb of Vibia is especially interesting in this regard, since there we find Mercury depicted at a single site, both with the second rod (in the judgement scene), and in his ordinary iconography with only his caduceus, in the image of him leading the chariot, where he is clearly in his role of psychopomp (Casagrande-Kim 2012, pp. 165–170). In the earliest example, the Jena vase, Mercury is depicted not only as a messenger of the gods, someone who acts upon others’ orders, but as an agent in his own right: as a true magician, who has the power to raise the dead and free their souls from the prison of the body. With his second rod Mercury acts for himself, makes decisions and creates: he has power over the dead in his own right; with the caduceus he is but a messenger who escorts them.
The Dalmatian examples are also found in funerary contexts, but what is missing is the iconographic surroundings: there is no sign of Hades or any other chthonic deity. We could argue that the image of Mercury is a shorthand, a symbol of a larger complex that is understood from the lone Mercury figure, maybe as the pictorial equivalent of the Orphic Gold Tablets that would serve to ensure a safe journey through the underworld to the ones who were introduced to that concept.
If that is the case, the source of the image, the entry point of the idea, may be seen in the Narona sarcophagus. The patron(s) of the sarcophagus had most likely ordered the images on the sarcophagus with a clear purpose in mind. The question is, can we attach the same meaning to all of the images in the province? The Dalmatian Mercury was clearly understood by a larger social group. These may have been initiates in the mysteries, as perhaps was the case for the burials with the Orphic Gold Tablets, but those were hidden from sight while Mercury is on full display. While it is common to assume that religion in antiquity was embedded in the social, political and cultural reality of the local community, we must also consider the wider picture, and what else might have been called to mind by that iconography. Although the depiction of Mercury holding a simple rod was unusual in the Roman Empire, the sight of a person holding a rod and using it for magic must have been relatively common.2 There are accounts which show that traveling magicians were frequent in the streets of the Roman towns, especially in the provinces (Tertullian, Apol. 23, Idol. 9, Praescript. haer. 43, Carn. Christ. 5.; Vett. Val. 2.17.57; Dickie, 2003, pp. 215–221). They would use magic formulae to make a person die in front of the crowd, then they would use wands bring them back to life, so that they could question them about the secrets they were privy to, once back from the realm of the dead. This kind of show was used as a marketing trick that would secure the magician a private hire to showcase their counselling, healing and divination skills, but it must have been an extremely memorable event. Magic was also commonly associated with Mercury, as it was with Orphism (Dickie, 2003, p. 195), especially in the context of Hermes/Mercury using a simple rod (Waele, 1927, p. 56).
The Orphic tradition might have been the vehicle that transported some of the more arcane symbolic knowledge of Hermes/Mercury across chronological, geographic and linguistic barriers to Roman Dalmatia. Might we picture Mercury as a ‘real’ magician, who would, in the Orphic tradition of Hermes Psychopompus, really re-awaken the deceased, unlike his human counterparts? Although this is appealing as an intellectual exercise, it is hard to be satisfied with that explanation for Dalmatian Mercury. We can trace some elements of his iconography across the longue durée, but we must wonder what information, and in what form, was available in second and third century CE Dalmatia. From the number and distribution of the finds we can conclude that this was a wider provincial phenomenon, not a localized one, and from the fact that they were made in local limestone, thus produced by local workshops, we can reject the proposition that this was a coincidence – that these were objects found in Dalmatia by chance.
It may be possible to propose an alternative interpretation by approaching the material from an anthropological viewpoint: would any of the ideas sketched above have been understood or considered by the local population of Roman Dalmatia in the second and third centuries CE, and if so, to what extent? In order to answer this question, it is necessary to investigate the cognitive constraints and possibilities of concept transmission.
The cognitive theory of religion uses research into higher mental functions, such as categorization, learning, memory, and emotion, to interpret the way religious concepts have been created, transmitted and retained (Andresen, 2001; Atran, 2004;Barrett, 1997; Boyer, 1994; Boyer, 2001; Lawson & McCauley, 1993; Pyysiäinen, 2001). It can be used to create a filter to rule out improbable hypotheses about ancient societies and their religion, especially in the case of ‘fringe’ phenomena including cultural appropriation, syncretism and the complex interactions taking place in the provinces.
The question of the spread of Roman state religion to the provinces is not an easy one. As Ando has pointed out, there were certain fundamental aspects of Roman state religion that might have made it difficult to transfer Roman religious practices to the provinces. For example, Roman religious practices were firmly embedded in particular sacred spaces (Ando, 2011, p. 442); Roman state religion also lacked firm religious leadership (Beard, North & Price, 1998, p. 21); had no real tendency towards proselytising (Price, 2012, p. 9), and no religious orthodoxy (King, 2003, p. 284). Roman gods were never stable, immutable concepts or timeless entities; they changed as the society that created them changed (van Andriaga, 2011, p. 135). Then there is the issue of cognitive constraints. Any culturally unfamiliar story will be distorted in transmission; with the retelling of the story, some culturally unfamiliar items are dropped, others replaced by the familiar ones, thus making the unfamiliar parts of the story less likely to be retained (Atran, 2004, p. 10). Boyer explains how there is always a new variant of any religious concept created through the reconstructing, distorting and changing of information in the process of communication (Boyer, 2001, p. 140). People do not invent gods; they infer them from the available information.
All of these elements are corresponding to the imagistic mode of religiosity. I use that term in reference to the seminal work in cognitive theory of religion, that of H. Whitehouse (Whitehouse, 2002), whose notions of doctrinal and imagistic modes of religiosity illuminate some of the apparent inconsistencies in research into Roman religion. The basic assumption of Whitehouse’s theory is that there exist two basic modes of religiosity, which he calls doctrinal and imagistic, and that they can be described through binary oppositions. The criteria that Whitehouse uses for this division are memory and the way in which religious concepts are transmitted, while the variables he introduces follow the psychological and sociopolitical axes. The central features of the imagistic mode of religiosity are a low frequency of rituals and a high level of excitement. It basic memory system is that of episode (or flash-bulb) memory: people easily remember a single ritual event in detail. In doctrinal religions, on the other hand, people use schemata and implicit scripts, while memories of several different events often tend to be blurred into a single record. The meaning of the ritual is not explained through teachings; on the contrary, knowledge about the ritual is created internally, by participants. In a religion which operates in an imagistic mode, the elements Western scholars are used to, since we have been culturally indoctrinated into the doctrinal mode (assuming the majority of Western scholars are primarily familiar with the Abrahamic religions of Judaism, Christianity and Islam), do not function. There is no orthodoxy, or clear transmission of religious knowledge as expected in doctrinal religions; instead knowledge is generated by internal revelation (cf. Burkert, 1987 p. 69). An image, in that system, has a particular role – it serves to make a specific cultural concept more easily remembered and transmitted, or even to take on a part of the cognition itself (Clark, 2008; Clark & Chalmers, 1998; Day, 2004). From the perspective of the cognitive theory of religion, the explanation of a cultural fact (or its distribution) should not be sought in a global macro-mechanism but in the combination of a series of individual micro-mechanisms. The questions we should ask are: what are the factors that would entice a public communication of an idea? What ideas will the audience construct from it? What transformation will take place in the cognitive system of the recipient? What are the factors that will ensure further distribution? Which characteristics should a concept have to stay stable (Sperber, 1996, p. 57)?
The main difference between the Dalmatian examples and those from other parts of the Empire is in the surrounding images: while we have images of the rest of the underworld in all of the other monuments, in Dalmatia Mercury is alone, outside of any narrative context (allowing for the fact that some of the examples from Dalmatia are preserved in a fragmentary context, so we do not know details of the images which may originally have surrounded them). In the context of cognitive theory, what we see here can be understood as the turning of the image of Mercury into a floating signifier that settled on a different point in the network.
It is quite common to see a concept which originally had a multi-layered religious meaning transformed into a magical symbol when it is transferred into a new cultural context. The concept of a deity is dispersed through the cultural surroundings (political, economic, and political) and the physical surroundings (images, architecture, texts), and added to the basic cognitive images already present in individuals’ minds. In the terms of cognitive network theory, the concept of Mercury is weighted down in the computational network by the corresponding concepts of cultural institutions: the weights are responsible for the activation of the concept in different circumstances, and the activation reinforces the existing connections (Davies, 1989; Fodor & Pylyshyn, 1988). If the activation of a nod in the connectionist network does not activate other concepts (if the sight of Mercury stops activating the complex Orphic idea of the immortal soul, for example, since it is not supported by its original surroundings), it is left ‘floating’, and can be anchored in other parts of the network. In other words, if we take a religious concept that is embedded in a specific cultural context away from that context, it will lose its full meaning, and it will be transformed into something else. This is rarely the case with doctrinal religions, such as Christianity, because they presume the existence of an independent field of religion, and have a related religious identity separate from other identities. This makes them successful in spreading all over the world. But religious identity in imagistic religions of antiquity cannot be understood outside of someone’s political and social role in the network. In that case, research into the ‘original’ meaning of a certain iconographic depiction often will not help with understanding the place that concept has in a new system. The local variant of Mercury in Roman Dalmatia was used in a funerary context, but outside of the usual narrative, and as an addition to the main theme of the monument, not as a central motif. In this sense we could talk about the image of Mercury with the second wand as a magical symbol, keeping in mind how difficult it is to draw a clear line between magic and religion, and the heavy load of that discussion (Frankfurter, 2002, p. 159).
The theme of Mercury Psychopompus is a part of an elaborate conceptualisation of the journey to the afterlife, one that is often recognized as Orphic, and the images of Mercury holding a second rod all belong to that rich symbolic background. The two rods that Mercury is holding are making him simultaneously a messenger of the gods, the connection between the worlds and a magician, an active agent who is capable of waking the dead, not only of guiding souls. But the question is what does the search for iconology in the Warburgian sense bring to the table if we are interested in the image from an anthropological point of view? If we search for the ‘original’ god, we could, for example, follow Frothingham all the way to Babylon where he finds Hermes as a deity of spring and fertilization, a messenger of the Great Mother (Frothingham, 1916, p. 175), or Kerenyi in his meditations on Jungian archetypes of the deity (Kerenyi, 1987). These are fascinating, but offer little in the way of answers to particular questions – such as, why did such an image appear in a certain time and place, and what did it represent to the people who decided to put it on their final resting places?
Exploring these images from an anthropological point of view may also help to resolve some issues presented by the material itself. Why there are many relatively simple images of this type in Dalmatia as opposed to relatively few, but exponentially more complex ones in Rome? Even if the image of Mercury as a divine Orphic necromancer entered Roman Dalmatia, it most likely lost the complex symbolic meaning it had in Italy or Greece, which could not be supported without the cognitive underpinning of a wider system, and was left as a magical amulet, perhaps contaminated with the image of traveling magicians. The fact that Mercury was depicted in isolation supports that explanation; we do not see complex narrative structures in these images, like the ones on the contemporary Hecate from Dacia, or the patients of his magical actions, like we have on the Jena vase.
I propose this as a small building block in the attempt to answer a greater question, relating to how we conceptualize religion in the provinces. Scholarly discussion in this area so far has mostly concentrated on identity and the extent to which religious concepts were accepted or rejected; I propose, following the footsteps of Ando and Rüpke, to understand religion in provinces as a separate sub-system of the wider system of religion in the Empire. Religious concepts, as they enter that system, do so from a different social and cultural system, and their character changes through their incorporation into a new network. What I propose is a dynamic conceptualization of religion as a network of different agents and the parts of the environment through which religious concepts are distributed, including images, epigraphy, rituals, sacred spaces, and architecture. New religious concepts in the provinces will be communicated in a different network, and that process will intrinsically change the concept, even if there is an attempt at full mimesis. Any conscious attempt to negotiate identities will be made through this dynamic network. That was even more the case, since the main mode of religiosity in the ancient world was imagistic, which is intrinsically much more susceptible to change and multimodality. The idea of religion as a network in constant negotiation and dialectical change stemming from its inner conflicts can take us beyond identities as a primary focus for the study of religious change.
1 Sonoc notes that there is a connection between the Aercura and Hecate, especially through the Istria and Dalmatia, but that is rather hard to prove. Although there is one potential image of Hecate in Dalmatia, there is no direct connection to the images of Mercury (Sonoc, 2006).
2 Some accounts for the use of staff in context of magic may be found in Papyri Graecae Magicae, for example PGM IV.2006–2125 (Here is the figure drawn on the skin: A humanoid figure, with the head of a lion, wearing a belt, brandishing a staff in the right hand, on which there is to be a snake.) and PGM I.262–347 (Put on a prophet’s robe, hold an ebony staff in your left hand and the phylactery in your right one, that is, the laurel twig). Translation by Ogden, 2002.
Abramić, M. (1932) ‘Spomenici iz bedema stare Salone’ Bulletino di archeologia e storia dalmata, 50, pp. 56–62.
Alderink, L.J. (1981) Creation and salvation in ancient Orphism [University Park, PA], Chico, American Philological Association, distributed by Scholars Press.
Ando, C. (2011) ‘Exporting Roman Religion’ in J. Rüpke (ed) A Companion to Roman Religion. Hoboken, NJ: John Wiley & Sons.
Andresen, J. (ed) (2001) Religion in Mind. Cognitive Perspectives on Religious Belief, Ritual, and Experience. Cambridge, UK, and New York, Cambridge University Press.
Atran, S. (2004) In Gods We Trust: The Evolutionary Landscape of Religion, Oxford and New York, Oxford University Press.
Barr-Sharrar, B. (1990) ‘How important is provenance? Archaeological and stylistic questions in the attribution of ancient bronzes’ in J. Walsh (ed) Small Bronze Sculpture from the Ancient World. Papers Delivered at a Symposium, Malibu, CA, The Museum, pp. 209–238.
Barrett, J.C. (1997) ‘Romanization: a critical comment’ in D. Mattingly and S.E. Alcock (eds) Dialogues in Roman Imperialism. Power, Discourse, and Discrepant Experience in the Roman Empire, Portsmouth, RI, JRA, pp. 51–64.
Beard, M., North, J.A., and Price S.R.F. (1998) Religions of Rome, Cambridge, Cambridge University Press.
Benoît, F. (1959) Mars et Mercure, nouvelles recherches sur l’interprétation gauloise des divinités romaines, Paris, Éditions Ophrys.
Bernabé Pajares, A. and Jimenez San Cristóbal, A.I. (2008) Instructions for the Netherworld: The Orphic Gold Tablets, Leiden, Netherlands, and Boston, MA, Brill.
Bernabé Pajares, A.J. (2008) Instructions for the Netherworld the Orphic Gold Tablets, Leiden, Netherlands, and Boston, MA, Brill.
Boyer, P. (1992) ‘Explaining Religious Ideas: Elements of a Cognitive Approach’, Numen, 39 (1), pp. 27–57.
Boyer, P. (1994) ‘Cognitive constraints on cultural representations: Natural ontologies and religious ideas’ in L. Hirschfield and S. A. Gelma (eds) Mapping the Mind: Domain Specificity in Cognition and Culture. Cambridge, Cambridge University Press, pp. 391–411.
Boyer, P. (2001) Religion Explained. The Evolutionary Origins of Religious Thought, New York, Basic Books.
Bremmer, J.N. (2002) The Rise and Fall of the Afterlife, London, New York, Routledge.
Bulić, F. (1885) ‘Iscrizioni inedite’, Bulletino di archeologia e storia dalmata, 8, pp. 41–45.
Burkert, W. (1987) Ancient Mystery Cults. Cambridge, MA, Harvard University Press.
Cambi, N. (1960) ‘Personifikacije godišnjih doba na spomenicima u Saloni’, Vjesnik za arheologiju i historiju dalmatinsku, 62, pp. 55–78.
Cambi, N. (1975) Sarkofazi na istočnoj jadranskoj obali (III–VII st.n.e.). PhD dissertation, University of Zagreb.
Cambi, N. (1980) ‘Antička Narona – urbanistička topografija i kulturni profil grada’ in A. Basler (ed) Dolina rijeke Neretve od prethistorije do ranog srednjeg vijeka, Split, Hrvatsko arheološko društvo, pp. 127–160.
Cambi, N. (1988) Atički sarkofazi u Dalmaciji, Split, Književni krug.
Cambi, N. (2010) Sarkofazi lokalne produkcije u rimskoj Dalmaciji (od II. do IV stoljeća). Split, Književni krug.
Casagrande-Kim, R. (2012) The Journey to the Underworld: Topography, Landscape, and Divine Inhabitants of the Roman Hades, PhD dissertation, Columbia University.
Clark, A. (2008) Supersizing the Mind. Embodiment, Action, and Cognitive Extension, Oxford and New York, Oxford University Press.
Clark, A. and Chalmers, D. (1998) ‘The Extended Mind’ Analysis, 58 (1), pp. 7–19.
Combet-Farnoux, B. (1980), Mercure romain. Le culte public de Mercure et la fonction mercantile à Rome de la république archaïque à l’époque augustéenne, Rome, Ecole française de Rome.
Cumont, F.V.M. (1922 [2005]) After life in Roman paganism. Lectures delivered at Yale University on the Silliman foundation. Eugene, Wipf & Stock.
Davies, M. (1989) ‘Connectionism, modularity, and tacit knowledge’, The British Journal for the Philosophy of Science, 40 (4), pp. 541–555.
Day, M. (2004) ‘Religion, off-line cognition and the extended mind’, Journal of Cognition and Culture, 4 (1), pp. 101–121.
Dickie, M. (2003) Magic and magicians in the Greco-Roman world, London and New York, Routledge.
Dütschke, H. (1882) Antike Bildwerke in Oberitalien. Leipzig, Engelmann.
Edmonds, R.G. (2004) Myths of the Underworld Journey: Plato, Aristophanes, and the ‘Orphic’ Gold Tablets, Cambridge, Cambridge University Press.
Edmonds, R.G. (2013) Redefining Ancient Orphism: A study in Greek Religion, Cambridge, Cambridge University Press.
Ferrua, A. (1971) ‘La catacomba di Vibia’, Rivista di Archeologia Cristiana, 47, pp. 7–62.
Fodor, J.A. and Pylyshyn, Z.W. (1988) ‘Connectionism and cognitive architecture: A critical analysis’, Cognition, 28 (1), pp. 3–71.
Frankfurter, D. (2002) ‘Dynamics of Ritual Expertise in Antiquity and Beyond: Towards a New Taxonomy of “Magicians” ’ in P.A. Mirecki and M.W. Meyer (eds) Magic and Ritual in the Ancient world, Brill, pp. 159–178.
Frothingham, A.L. (1916) ‘Babylonian Origin of Hermes the Snake-God, and of the Caduceus’, American Journal of Archaeology, 20 (2), pp. 175–211.
Gramatopol, M. (1982) Dacia antiqua. Perspective de istoria artei si teoria culturii, Bucuresti, Meridiane.
Gramatopol, M. (2000) Arta romană în România, Bucuresti, Meridiane.
Gulizio, J. (2000) ‘Hermes and e-ma-a2. The Continuity of his Cult from the Bronze Age to the Historical Period’, Živa Antika, 50, pp. 105–116.
Herrero de Jauregui, M. (2010) Orphism and Christianity in Late Antiquity, Berlin and New York, Walter de Gruyter.
Hirschfeld, O. and Schneider, R. (1885) ‘Bericht uber eine Reise in Dalmatien’, Archaeologisch-Epigraphische Mittheilungen aus Oesterreic.Ungarn 9, pp. 1–84.
Imamović, E. (1977) Antički kultni i votivni spomenici na području Bosne i Hercegovine. Sarajevo, Veselin Masleša.
Jastrzębowska, E. (1979) ‘Les scenes de banquet dans les peitures et sculptures chrétiennes des IIIe et IVe siècles’, Recherches Augustiniennes, 14, pp. 3–90.
Jongman, W. (2007) ‘The Early Roman Empire: Consumption’ in W. Scheidel, I. Morris and R.P. Saller (eds) The Cambridge Economic History of the Greco-Roman world, Cambridge, Cambridge University Press.
Kerenyi, K. (1987) Hermes, Guide of Souls. The Mythologem of the Masculine Source of Life. [3rd ed.] Dallas, Spring Publications.
King, C. (2003) ‘The Organization of Roman Religious Beliefs’, Classcal Antiquity, 22 (2), pp. 275–312.
Koch, G. (1933) Sarkophage der römischen Kaiserzeit, Darmstadt, Wissenschaftliche Buchgesellschaft.
Lawrence, M. (1932) ‘Columnar Sarcophagi in the Latin West: Ateliers, Chronology, Style’, The Art Bulletin, 14 (2), pp. 103–165.
Lawrence, M. (1958) ‘Season Sarcophagi of Architectural Type’, American Journal of Archaeology, 62 (3), pp. 273–295.
Lawson, E. and McCauley, R. (1993) Rethinking Religion: Connecting Cognition and Culture, Cambridge, Cambridge University Press.
Lulić, J. (2015) Aspekti religijske skulpture na području rimske provincije Dalmacije. PhD dissertation, University of Zagreb.
Lulić, J. (2016) ‘Illyrian Religion and Nation as Zero Institution’, Studies in Visual Arts and Communication – An International Journal, 3 (1), online.
Ogden, D. (2002) Magic, Witchcraft, and Ghosts in the Greek and Roman Worlds: A Sourcebook. Oxford, Oxford University Press.
Patsch, K. (1894) ‘Rimske starine iz Goražda’ Glasnik Zemaljskog muzeja u Sarajevu, 6 (1), pp. 53–56.
Patsch, K. (1899) ‘Arheološko-epigrafska istraživanja iz rimske provincije Dalmacije’ Glasnik Zemaljskog muzeja u Sarajevu 11 (1), pp. 69–123.
Patsch, K. (1900) ‘Archäologich-epigrapische Untersuchungen zur Gesichte der römanischen Provinz Dalmatien IV’, Wissenschaftlische Mittheilungen aus Bosnien und der Hercegovine, pp. 33–166.
Patsch, K. (1914) ‘Zbirke rimskih i grčkih starina u bos.-herc. zemaljskom muzeju’, Glasnik Zemaljskog muzeja za Bosnu i Hercegovinu, 26 (1), pp. 141–221.
Petersen, E. (1881) ‘Der Reliefschmuck der Hekate von Hermannstadt’, Archäologischepigraphische Mitteilungen aus Österreich-Ungarn, 5, pp. 193–202.
Platt, V. (2011) Facing the Gods. Epiphany and Representation in Graeco-Roman Art, Literature and Religion, Cambridge, Cambridge University Press.
Price, S. (2012) ‘Religious Mobility in the Roman Empire’, Journal of Roman Studies, 102, pp. 1–19.
Pyysiäinen, I. (2001) ‘Cognition, Emotion, and Religious Experience’ in J. Andresen (ed.) Religion in Mind. Cognitive Perspectives on Religious Belief, Ritual, and Experience. Cambridge and New York, Cambridge University Press, pp. 70–93.
Russell, B. (2010) ‘The Roman Sarcophagus “Industry”: A Reconsidiration’ in J. Elsner and J. Huskinson (eds) Life, Death and Representation: Some New Work on Roman Sarcophagi, Berlin, De Gruyter.
Sanader, M. (1986) Kerber u antičkoj umjetnosti, Split, Logos.
Sanader, M. (1994) ‘O kultu Herkula u Hrvatskoj’ Opuscula Archaeologica, 18, pp. 87–114.
Schadow, P. (1897) Eine attische grablekythos, Jena, G. Neuenhahn.
Sonoc, A.G. (2006) ‘Erz und Schlacke in provinzialrömischen Gräber aus Dakien’ Acta Terrae Septemcastrensis Journal, Proceedings of the 7th International Colloquium of Funerary Archaeology, Special number (Bibliotheca Septemcastrensis, XVII), pp. 209–222.
Sperber, D. (1996) Explaining Culture. A Naturalistic Approach, Oxford and Cambridge, Blackwell.
Toynbee, J.M.C. (1996) Death and Burial in the Roman World, Baltimore, Johns Hopkins University Press.
Waele, F.J.M. (1927) The Magic Staff or Rod in Græco-Italian Antiquity, Gent, Drukkerij Erasmus.
van Andriaga, W. (2011) ‘New Combinations and New Statuses. The Indigenous Gods in the Pantheons of the Cities of Roman Gaul’ in J.A. North and S.R.F Price (eds) The Religious History of the Roman Empire. Pagans, Jews, and Christians, Oxford and New York, Oxford University Press, pp. 110–136.
Whitehouse, H. (2002) ‘Modes of Religiosity: Towards a Cognitive Explanation of the Sociopolitical Dynamics of Religion’ Method & Theory in the Study of Religion, 14, pp. 293–315.
Whitemarsh, T. (ed.) (2010) Local knowledge and Microidentities in the Imperial Greek World, Cambridge, Cambridge University Press.