Paper Information: Title: Whatever Happened to the Heroes? Ancestral Cults and the Enigma of Romano-Celtic Temples Author: Colin Forcey Pages: 87–98 DOI: http://doi.org/10.16995/TRAC1997_87_98 Publication Date: 16 April 1998 Volume Information: Forcey, C., Hawthorne, J., and Witcher, R. (eds.) (1998) TRAC 97: Proceedings of the Seventh Annual Theoretical Roman Archaeology Conference, Nottingham 1997. Oxford: Oxbow Books. Copyright and Hardcopy Editions: The following paper was originally published in print format by Oxbow Books for TRAC. Hard copy editions of this volume may still be available, and can be purchased direct from Oxbow at http://www.oxbowbooks.com. TRAC has now made this paper available as Open Access through an agreement with the publisher. Copyright remains with TRAC and the individual author(s), and all use or quotation of this paper and/or its contents must be acknowledged. This paper was released in digital Open Access format in April 2013. Whatever Happened to the Heroes? Ancestral cults and the enigma of Romano-Celtic temples by Colin Forcey Int roducti on In this paper I wish to re-interpret the nature and 'purpose' of Romano-Celtic temples in first century AD Britain, in the light of fairly widespread indications of elite mortuary use in and around them, and their related association with ritual pits/wells. Such an association is suggestive of connections to a mythical Underworid: the realm of the dead (the ancestors) and concomitantly with the memory of the past. This persuades me to controvert an a-temporal, functionalist classification of the phenomenon of Romano-Celtic temples (Rodwell 1980 lists twelve 'types') in favour of an historical/genealogical (Dean 1994; Foucault 1991; Nietzsche 1996 [1887] :57-58) approach to their interpretation, relating them to native memories of an 'heroic' past, before the Roman Conquest. The key site, which has been the stimulus for my re-interpretation of Romano-Celtic temples in the early Roman period, is the mortuary enclosure at Folly Lane, excavated by Rosalind Niblett (Denison 1994:5; Niblett 1992, 1995). This enclosure is situated on a hill to the north of Verulamium overlooking the city. In the immediate post-Roman Conquest period it was the locus of a complex multi-stage funerary ritual ll1volving the laying out, cremation and deposition of an important individual of the native elite (these rites are similar to those observed at the recently excavated, first century mortuary enclosures at Stanway outside Colchester (Crulluny 1997 23-27». Some fifty years after the Roman Conquest (in the Flavian period), a Romano-Celtic temple was built within the enclosure at Folly Lane, directly on top of the area where the body was cremated, seemingly as an act of commemoration. Though it is tempting to suggest that such a temple is in fact a mausoleum which have been mis cJassified, the evidence of archaeology (see below) suggests that religious and mortuary use of Romano Celtic temples was not mutually exclusive and that it is our classificatory system itself which is at fault. Buildings recognised as mausolea (Black 1986; Jessup 1959; Toynbee 1971;) and those recognised as Romano-Celtic temples (Lewis 1966; Rodwell 1980) often have exactly the same form. (Recently the portmanteau term temple-mausoleum has come into use for (un)certain structures (Williams and Zeepvat 1994». Similarly the rectilinear plan and cardinal point orientation of certain supposed religious precincts: temene (such as those enclosing the Hayling Island temple (Downey, King and Soffe 1980; King and Soffe 1994), the Gosbecks temple (Hawkes and Crummy 1995 95-105) and the Thetford ritual enclosure (Gregory 1992) demonstrate a very striking resemblance to mortuary enclosures, such as the recently excavated examples at Stanway (Crummy 1997:23-27). If we discard our own modern cultural preconceptions, which divide the archaeology of death and burial from that of religion, It is possible to fe-interpret the nature of such structures including the possibility of a combination of mO!1uary and religious rites. Such a combination, in the form of ancestral and heroic cults, are not unexampled in the classical world, and are found in other provinces of the Roman Empire at this time, most notably Greece (Rohde 1925: 115-216; see Pausanias' Guide to Greece for an eyewitness account of these cults in the second century AD). Ancestor shrines and cults are well known in modern ethnographic contexts (e.g. India, Sri vinas I 952160--63; China, Ahern J 973) The political implications of such an ancestral or 'heroic' cult on the hill above Verulamium also bears some examination, in terms of the organisation of space and the hegemony of Rome. Was the memory of the person buried on the hill a challenge to the dominant order as manifested by the 88 Colin Forcey 'Romanization' of the city in the valley below? (Branigan 1973; Forcey 1997; Haselgrove and Millett 1997 :286; Hunn 1992). Some of the activities in its vicinity certainly look challenging. Adjacent to the enclosure were some fifty-two ritual shafts, in one of which a skull with traces of grisly mutilations was discovered. Marks on the skull suggest that it had been, literally, defaced. This horror was, if anything, compounded by the nearby find of a face pot on which a similar mutilation had been practised: the face on the pot had been cut out and deposited in the same pit. A number of other face pots were found in the other shans (Denison 19945; Mays and Steele 1996, Niblett pers. conun.). Employing Hodder's idea of an 'archaeology of context', rather than arbitrarily dividing phenomena on the basis of twentieth century functional or typological categories, I will investigate the general significance of the constellation of phenomena seen at Folly Lane. I will discuss the significance of the ritual pits found in the vicinity of this and other Romano-Celtic temples in terms of beliefs about the Underworld (the realm of the dead). I will also try to make some sense out of the seemingly strange finds from such pits. In all this, I am deeply unimpressed by any supposed methodological/epistemological impediment to our understanding of ritual/religion, as articulated most explicitly in Hawkes' (1954) influential 'ladder of inference'. According to this conunonly held view (defended by Trigger 1989:266, 327, 392-5) the domain of ritual/religion is conceived of as a deeply mysterious realm of bizarre happenings and inaccessible entities, only accessible to the archaeologist through a strange form of inference from the supposedly readily apparent fields of technology, through the increasingly obscure stages of economy and socio-political organisation. It is apparent, however, that such amateur 'philosophising' seems mainly to involve an uncritical fetishism of modern western cultural categories on the part of the archaeologist, inclusive of a post-Protestant theological hangover which equates ritual with 'the bizarre' (read 'egregious Popish impostures') and religion with 'belie!" in spiritual beings' (read: 'justification by faith'). This is less than inadequate, and ignores the sociological direction the study of ritual and religion has taken since Durkheim's seminal work of the early years of this century (Durkheim 1915). The next three sections will address the implications of Folly Lane for a re-assessment of other examples of Romano-Celtic temples, highlighting: 1. the connection of Romano-Celtic temples with the remains of the dead. 2. the related connection with ritual pits, which have chthonicfUnderworid associations. 3. the place of Romano-Celtic temples as the focus of memory and the negotiation of power in the early Romdn period. Mortuary and votive rites Indications of funerary activity at or around Romano-Celtic temples are surprisingly widespread. Unfortunately the prejudgement of a structure as a 'temple' (conceived on the model of a classical temple) has often led to the dismissal of (usually major) anomalies in the evidence (e.g. the presence ot" human bones, burnt areas, burnt and broken items of personal adornment, the location of the temple in the middle of a cemetery etc. etc.) which do not support the initial formulation. Taphonomic considerations, inadequate excavation and recording, and the sheer complexity of funerary ritual of this period have also hampered the recognition of LPRIA (Late Pre-Roman Iron Age) and early Roman pyre sites and cremation burials in Britain (see Perch, this volume). The recognition of funerary activity at Romano-Celtic temple sites would help to fill such lacunae in our evidence. The confusion has been increased by a seeming inability to distinguish between grave goods and votive offerings. It is frequently alleged tht damaged goods (supposed votive offerings) found at Romano-Celtic temples have been 'ritually ki'.led' in order to facilitate their entrance into the Other World; this in strange contrast to the univcr· 'I I lctice of the classical world where votive offerings, as gifts to the gods, were never broken (5. Burke pers. comm.)[IJ. Where deliberately broken objects occur on a site we should Whatever Happened to the Heroes? 89 be wary of the hand-me-down mystical metaphysics of 'ritual killing'. The deposition of such ohjects is much more redolent of funerary rites: their broken nature explained either as the result of their being subject to the names of a cremation or as the result of a performative rite symboLising the breaking of the body of the deceased as a rite of passage from the world of the living to the world of the dead (noted 1I1 ethnographic contexts: HeItz 1960; Metcalf and Huntingdon 1979; Srivinas 1952, Van Gennep 1960) The respective nature of grave goods and votive offerings and the circumstances of their deposition differ to a certain extent and some crude distinctions can be made. Grave goods typically include items of personal adornment, pottery vessels (often containing consumables) and chariot equipment. Typical votive offerings include figurines of gods, miniature weapons, and cast metal letters to write dedications. Coins figure in both categories. Context is all important. The deposition of several items of (often broken and burnt) personal adornment and pottery etc. of the same age in a single context would be a typical signature of grave goods and mortuary ritual. Votive offerings by contrast would typically be placed in the temple over a considerable period of time, sometimes centuries, and would very rarely make into the archaeological record at all. Such items were not buried in the first instance (they were placed within the temple as gifts to the gods) and notoriously were subject to appropriation by emperors, adventurers, thieves, etc. (for such depredations see Ulpian Duty of the Proconsul, Book 7, in Digest I.J 8.13, quoted in Goodman 1997: (45). Using the above crude criteria of evidence, a mortuary signature comprising the presence of human bones comhined with burnt areas, pyre debris and/or broken grave goods buried in a single act of deposition can be discerned at such prominent Romano-Celtic temples as Harlow (the original type site of a Romano-Celtic temple in Britain) (Bartlett 1988; Burnham and Wacher 1990: 183- 188; France and Gobel 1985; Frere 1940:161-171;); Lancing (Bedwin 1981; Winbolt 1935:59-60); Hayling Island (Downey, King and Soffe 1980; King and Soffe 1994) Greenwich, London (Merrifield 1983:139; Sheldon and Yule 1979) and Cosgrove, Northants. (Quinnell 1991). Location of temples provide further clues: the temple at Lancing is the focal point of a Roman cemetery (Bedwin 1981; Frere 1940: I 61- 171; Winbolt 1935:59-60); Bourton Grounds temple, Buckinghamshire is adjacent to the Thornborough Roman barrows (Lewis 1966); Weycock hill temple. Berkshire is adjacent to a cemetery (Cotton 1956); Colchester 'temple 6' is next to a walled cemetery and temples '7', '10' and 'A' are within the extra-mural cemetery zone outside the West (Balkerne) Gale of the city (CrLIIIDllY 1980) (see further below, for a re-interpretation of temples 2-5 at Sheepen, Colchesler). The recently excavated (concentric double square) structure at Bancroft (Buckinghamshire) has both mortuary features (many human bones from an original vault) and votive indications and has been designated by its excavators as a temple-mausoleum (Williams and Zeepvat 1994). A similar structure has been recorded at Ratham Mill, Funtington. West Sussex (King and Soffe 1983) The funerary dssociations of some structures are Indeed so pronounced lhat they have been assigned to an entirely different category as mausolea. Such 'mausolea' (e.g. Harpendon, Herts. (Lowther 1937); Welwyn, Herts. (Rook, Walker and Denston 1984), Wood Lane End, Herts. (Neal 1983. 1984), and Lullingstone, Kent (Meates 1979, Rodwell 1980» usually have exactly the same form as Romano-Celtic temples. The decision to allocate to either category seems to be based on the (mis)chances of archaeological retrieval and can be quite arbitrary (e.g. at the Springhead temple complex 111 Kent, one concentric double square structures (erected over four Infant burials - one under each corner) is designated a 'temple', whilst another nearby (which encloses two infant burials in lead coffins) is labelled a 'mausoleum' (Burnham and Wac her 1990: 192-198». I bel ieve that the posited similarity of form of Romano-Celtic temples and mausolea is in [act an identity and that supposed category distinction are an artefact of our own cultural consciousness where the discourses of death and religion are distinct academic specializations (Lewis 1966 on temples versus Toynbee 1971 on mausolea) Though the limitations of this approach have 90 Colin Forcey become increasingly apparent, and the neologism 'temple-mausoleum' has been coined (Williams and Zeepvat 1994), the full implications of the problem have yet to be grasped. Our difficulty appears to be a conceptual (even 'theoretical') one. The possibility of some kind of ancestral or heroic cult, with a combination of votive and mortuary rites might go a long way to explain the many odd features of Romano-Celtic temples. For instance the Muntham Court Romano-Celtic temple in Sussex has a hearth in the middle of the cella where one would presumably (on the model of classical temples) have expected the votive offerings to be placed (votive offerings were found, outside the temple). Nearby was a 200 foot deep ritual pjt/well with dog skeletons at the bottom: of chthoniclUnderworid significance (Holleyman and Burstow 1955; Burstow and Holleyman 1957; Green 1976:220; Black 1986). It is to the chthoniclUnderworld significance of ritual pits in relation to Romano-Celtic temples that I now turn. Rilual pits/wells, the Underworld and the god MercUlY Ritual pits/wells are often found in the vicinity of Romano-Celtic temples [2]. Such ritual pits/wells have long been regarded as quintessentially 'Celtic' and linked to springs, rivers and bogs as repositories of offerings/sacrifices (Cunliffe 1988:359; Ross 1968). However, many examples are in fact known cross-culturally, including from the classical world (Webster 1997). Probably the best known example of this is the ritual shaft (Mundlls) on the Palatine Hill in Rome, which was regarded as giving access to the Underworld and from which the spirits of the dead (the di manes: see below) were supposed to issue on a certain three days of the year (Webster 1997: 139; for classical conceptions of the nature of the Underworld see Bailey 1935; Cumont 1959; Onians 1951: 254). Ritual pits/wells have been found within or adjacent to the temene of Romano-Celtic temples at Great Dunmow and Great Chesterford in Essex (Ross 1968:264-265; Wait 1985:407--408), at Muntham Court in Sussex (see above), possibly at Southwark, London (Merrifield 1996), at Weycock Hill (Cotton 1956:55-56), and at the Lower Brook Street site in Winchester (Biddle 1975:298-299; Ross 1975). At the latter site an anthropomorphic wooden statuette of the female Celtic goddess Epona was found at the bottom of the pit, dated to 150-60 AD. The statuette holds a key in her right hand. Epona is most often associated with horses, though she was also a chthonic goddess, "guiding the souls of the departed to the otherworld of Celtic belief' (Ross 1975:336). The key symbolises "her power to unlock the doors of temple, stable or home; and of opening the gates of eternity for the departed" (Ross 1975:336). A related indication of Underworld/chthonic associations is provided by depictions and dedications to the god Mercury in association with Romano-Celtic temples, ritual pits and graves. Though commonly thought by archaeologists to be exclusively associated with trade and money (e.g. Hawkes and Crummy 1995:95, 101, 105) Mercury, in another aspect, as the Roman equivalent of the Greek Hermes Psychopompos, was the conductor of the souls of the dead LO the Underworld. The god Mercury was, from a very early period, completely assimilated to the Greek god Hermes in all his aspects (Bailey 1935: 117-118). Though the mortuary role of Mercury has not often been stressed in the context of Roman studies; literature, epigraphy and archaeology all attest to it. In Roman literature we can cite the works of Virgil and Horace (and Seneca's Apocolocyntosis). Here Mercury was not only the conductor of the souls of the dead to the Underworld but also acts to keep the dead in place: to prevent them transgressing the bounds of the world of the dead and returning to the terrestrial sphere (Horace Odes 1.10, 1.25; Scheid 1996:962). In this respect Mercury is what is technically called an 'infernal deity' or god of the Underworld, sharing this honour with such as Hecate, Persephone and Tartarus, and with the divinised spirits of the dead (all the above classed indiscriminately as the di manes)(Merrifield 1987: 141) Whatever Happened to the Heroes? 91 Roman sarcophagi from Gaul often show Mercury in this role (Esperandieu 1907-66) In both Britain and Gaul, Mercury has quite literally been found in an 'Underworld' context. At Poitiers a ritual pit/well (120 feet deep) adjacent to two Romano-Celtic temples produced a votive inscription to Mecurius Adsmerius (Ross 1968:278). At Emberton in Buckinghamshire an altar portraying a non classical Mercury was found at the bottom of a sacred well/ritual pit which is possibly associated with a shrine (Ross 1967:30, 192) As a god of the Underworld (part of the collective di manes who occupied the lower sphere in opposition to the celestial gods: the di superi), MercurylHermes was often invoked in curse tablets throughout the Roman Empire, in which we find victims typically devoted 'to the demons of the Underworld, Hermes of the Underworld, Hecate of the Underworld, to Pluto, Persephone, the Fates of the Underworld, to all the gods and Cerberus' (Tomlin 1988:62). Putative entrances to the Underworld, such as the sacred spring at Bath attracted such curse tablets, to Celtic and/or Roman Underworld deities (Tomlrn 1988). Such curse tablets invoking the aid of Mercury have been found in profusion at some Romano-Celtic temples, such as the many from the Uley temple in Gloucestershire (Woodward and Leach 1993; Tomlin 1993). At the Kelvedon Romano-Celtic temple in Essex, and at the Holbrooks ritual site (associated with the neighbouring Romano-Celtic temple at Harlow) curse tablets dedicated to Mercury have been found in associated ritual pits/wells (Conlon 1973; Tomlin 1988:61; Wait 1985:410; Wright 1958: I SO). The Bancroft temple-mausoleum and vi lIa complex has iconography (figurines and sculpture depicting cockerels) redolent of the cult of Mercury (Green 1974, 1994:321). Such have also been found in the cemetery around the Lancing temple (Frere 1940: 161-171; Winbolt 1935:59-60). As indicated above for Folly Lane, human skulls, face pots and masks are often found within ritual pits/wells (Ross 1962, 1968; Wait 1985:321-335) In addition to tile Folly Lane examples above there are examples from Holbrooks-Harlow where a grotesque face mask (which is not a typical theatrical mask, but bears close resemblance to the equally gruesome examples of unknown function found at the Altbachtal Romano-Celtic temple complex at Trier: Wightman 1970) has recently been found deposited in a Roman period well (Bartlett pers. comm.). At Bath the sacred spring produced another enigmatic Romano-Celtic tin mask, perforated with nail holes (Cunliffe 1988; see Dungworth, this volume, for further thoughts on the ritual aspects of nails). What this head/face/mask ritual signified is difficult to interpret. Some concern with the ancestors seems likely. It may, however indicate subversive magic (self-styled magicians and sorcerers were commonplace in the Roman world: Macmullen 1966) rather than being part of an official cult (or indeed 'anti-magic' - in modern Haiti decapitation of corpses and transposition of head and feet is practised to prevent sorcerers re-animating the corpse as a 'zombie'). Some form of head-cult and the ritual display of ancestor masks and deposition of skulls and statue heads is manifested in both the Celtic and classical world in both religious and magical contexts (Onians 195] :96-167,507-508; Ross 1962, 1967:127-167, 1968; Webster 1997) Ancestral masks were a common feature of Roman funerals (Pliny Natural History 35:2; Polybius Histories 6:53-54). It is of interest in this connection that the Latin word for 'ghost' and 'mask' are expressed by the same word: larva (Onians 1951:135). It may be relevant that the Romans had "the belief that the life-soul which survives death is in or is the head" (Onians 1951 :507). This is exemplified in such iconography as the 'Colchester Sphinx' (the Sphinx is a classic symbol of death, often found depicted on tombs) which holds a head, symbolising the soul of the deceased, between its paws (Crununy 1997: 108). Historians record that when a Roman general 'devoted' both himself and the enemy to the infernal gods (the di manes) - as a desperate expedient to win a close fought battle - it was his head which he offered them (Onians 195 I: 132). We have here a probable interweaving and complex re-interpretation of an intricate web of 'Celtic' and Roman beliefs and practices. Further research into the context of such finds would be productive, coupled with an awareness of the potential complexity of ritual/religious belief and practice (on the head cult see further: 92 Colin Forcey Braithwaite J 984; Cotton 1996; Marsh and West 1981; Merrifield 1987:46; Ross 1962, 1968; Wait 1985321-335). Memory and the negotiation of power There has been a presumption that Romano-Celtic temples functioned in a similar way to classical temples, with a continuous placement of votive offerings within the precinct, over any given period of time, including in some cases 'continuity' from the LPRIA through the Roman period (Wait 1985) However, we may instead be seeing the commemoration of events: discontinuities mediated by memory rather than continuous ritual performances (Williams, this volume). Such commemoration seems to be the case at Folly Lane (see above) and can be seen at other sites, such as the Hayling Island 'temple'. Here an LPRIA circular timber building with an associated dismembered inhumation and burnt and broken grave goods (a funeral event?) was replaced in the Flavian period by a stone shrine which mirrored its form. The coin sequence here points to a discontinuity between these two events (Haselgrove j 987). The lemenos around the shrine is similar in plan and orientation to those of the first century mortuary enclosures at Stanway, near Colchester. The constructional style and technique of the stone shrine at Hayling Island has links with the nearby palace of Cogidubnus at Fishbourne and is possibly linked to his dynasty (Downey, King and Soffe 1980; King and Soffe 1994) A site with similarities to both Hayling Island and Stanway is the ritual site at Fisons Way, Thetford (Gregory 1992). Here structures and graves dating back to the first century BC were given a massive final embellishment, in the mid first century AD, in the form of a series of concentric rectilinear enclosures with internal round huts. However the site was deliberately destroyed and levelled soon after, suggesting that its final embellishment was relateu io an event (the funeral of King Prasutagus'n rather than an on-going process. A much later temple near this site seems possible from the evidence of small finds: this structure has however possibly been destroyed by a nearby modern factory. Such discontinuities can be explained if we accept that monumentalization was mediated by memory rather than necessarily by continuous ritual performance. These memories could refer to myth as much as 'reality': the 'temples' at Harlow and Lancing are built adjacent to Bronze Age barrows, which probably had some legendary significance: the past was inscribed with meaning, which affected activity in the present (Williams this volume). At both these sites there is also evidence of preceding LPRIA structures and possible funerary activity, It may be relevant that the stone structures at Harlow and Lancing are built at the Flavian 'event horizon', when the last generation of those who witnessed the conquest would be dying out (Harlow: Bartlett 1988; Burnham and Wacher 1990:183-188; France and Gobel 1985; Lancing: Bedwin 1981; Frere 1940:161-171; Winbolt 1935:59-60), The problematic of continuity/discontinuity is also apparent at the four first century Romano-Celtic temples at Sheepen (west of Colchester) (Crummy 1997:86, 107), This temple complex seems to have no links with Iron Age predecessors. The area where they are situated was a thriving industrial, trading and manufacturing area before the Boudiccan revolt. After destruction, caused during the revolt or by Roman counter-insurgency afterwards, this whole area was occupied by a series of temple complexes. Though there is no continuity of function or ritual practice it is however possible that these temples also, related to the past through the operations of memory. Did they commemorate a dispossessed native aristocracy who fell victim to the Boudiccan revolt and its aftermath (Crummy 1997:86, 107). Both 'continuity' and 'discontinuity' need to be deconstructed. The Cult of King Cunobelinus at Cos becks ? Traditional functionalist explanations of Romano-Celtic temples have tended to regard them as being situated on tribal boundaries (Stevens 1940). This liminal position is supposedly conducive to Whatever Happened 10 the Heroes? 93 their operation as temple-rural fair complexes, with traders utilising such sites to exchange goods (Burnham and Wacher 1990:40-41). At Gosbecks (Crummy 1997:27-28, 102-107; Hawkes and Crummy 1995:95-105) the discovery of a bronze statue of Mercury, in a field north of the temple enclosure was adduced to support this interpretation: Mercury, being the god of traders (Hawkes and Crummy 1995:95, 101, 105). This explanation fails to convince, as I suspect that trading activities would have been more convenient in the forum of Colchester, two and a half miles away, and also because Mercury was not merely a god of trade but (as I have mentioned above), in another aspect, was the conductor of the souls of the dead to the Underworld. The statue of Mercury from Gosbecks may Indicate that the Romano-Cdtic temple here was dedicated to this god. Perhaps here, as elsewhere, he had a protective role with regard to the di manes - the divinised spirits of the dead. (A recent find at Gosbecks of an iron ring with an inset gemstone depicting the god Mars has led Crummy to suppose that this was the god worshipped at Gosbecks (Crummy 1997: 106-107). With respect to this I would suggest that in the 'inlerprelario Ramona' there is an equivocation between Mars and Mercury (as seen for instance at the Romano-Celtic temple at Uley (Woodward and Leach 1993» who both seem to represent the same Celtic Horned god figure (Ross 1967:134), and that this Celtic Horned god also has chthonic associations.) The ditch delimiting the inner enclosure at Gosbecks had a coin of Cunobelinus in the primary fill and seems to date from the immediate pre-Conquest phase. At a later period a Romano-Celtic temple was built within the corner of this enclosure. The placement of the temple in the corner has been regarded as odd and recourse to the sacred tree argument has been made to explain what was in the middle (Lewis 1966). If, however, we compare one of the nearby Stanway mortuary enclosures with the Gosbecks enclosure the similarities in plan, orientation and placement of the entrance are remarkable (compare plans in Hawkes and Crummy 1995). Also, the association of a Romano-Celtic temple with the Gosbecks enclosure is reminiscent of the Folly Lane mortuary enclosure. Perhaps the enclosure at Gosbecks is a mortuary enclosure as weJl? This would explain the odd placement of the temple, if rather than a tree there was a burial in the middle of the enclosure. The person buried there would obviously be of some importance, as we can see from the increasing elaboration of concentric outer enclosures to the east, and the placement of the temple In the corner. Nearby, to the south, is a large enclosure which Philip Crummy interprets as the original residence of King Cunohelinus (Crummy [997: 17-18). In later times a theatre was built in the immediate vicinity and a road connected the complex directly to the Colonia. It is plausible that all this activity was connected to the fame of the person buried within the enclosure. Who was this person? I suggest, with Phi Ii p Crummy (Crummy 1997:27-28, 102-107), that it was none other than Cunobel inus himself, and that his person was the object or focus of continuing devotion well into the Roman period. Such an Heroic cult 01 Cunobelinus would provide a native counter-weight to the cult of 'Claudius the God· at the Colonia. Though onc might wonder why the Romans would allow the veneration of a native leader it is probable that Cunobelinus was a fairly safe symbolic figure for them. Having died some two years before the Conquest he would not be an obvious symbol of resistance to Roman rule (unlike the renegade Caratacus). The memory of Cunobelinus could have been a locus In the negotiation of power between Roman and native. The Romans could legitimate their rule by claiming to follow the example of King CunobeiInus rather than their antagonist King Caratacus. In the classical context the combination of mortuary and sacred rites in the form of a sacrificial cult with special rites around the tomb of a Hero is common in the Greek-speaking part of the Roman Empire. These hero tombs often purported to be of the mythical founders of the city but also included historical personages (Rohde 1925: 115-216). The combination of mortuary and sacred rites with an ancestral cult was not completely unknown, also, in the western, Latin-speaking part of the empire (Weinstock 1971 :287-296), which was heavily intluenced by beliefs and practices from the Greek East and from Magna Craecia (southern Italy). In Roman myth and ritual practice the dead were assimilated to the infernal gods - the di manes 94 Colin Forcey (Bailey 1935:297-299; Onians 195 I :264-270; Radice 1973: 158). 'When a Roman burned his dead father, as soon as he came upon a bone, he said that he (the dead) had become a god' (Onians J9 5 1:267, quoting Yarro). According to the jurist Gaius, places where the dead were buried, and grave monuments - which in the early Roman period often had the form of altars - were literally the possession of the di manes (/nslilUles of Caius 2.2-9; Fraser and Nicholas 1958: 120). The ubiquitous dedicatory formula for tomb inscriptions of the imperial period is DIS MANIBUS. This, the Latin dative form of di manes, is virtually untranslatable into English, due to grammatical peculiarities and because of our different world-view and cosmology. However this Latin dedicatory formula combines notions of the shades of the dead (hopefuLLy beneficent ghosts) (manes) who inhabit the underworld (also 'manes'), and notions of the divine (di = gods) and bears the import: 'To the infernal [underworld] gods'. That is, the tomb is devoted to the divinised spirits of the dead and the infernal gods (both of these entities comprehended as one within the term dis manibus) who inhabit the Underworld in a similar way that temples and votive offerings were devoted to the celestial gods (dis superis) who were conceived as 'spirits in the sky' (Bailey 1935:297-299). The presence of an heroic cult in the Celtic world also, is hinted at by the Greek ethnographer Posidonius who uses the terminology of the heroic cult to describe Celtic temples in southern Gaul (Webster 1995:446-447) The cult of the head, in the classical and Celtic worJd, associated with Romano-Celtic temples (see above) possibly evinces a similar concern with ancestors. Conclusion Romano-Celtic temples were a locus in the negotiation of power between Roman and native. As such they were polysemic and could mean different things to different groups in society at different times. On the Roman side the appropriation of the authority of pre-Conquest rulers, invested in such monuments, could provide some semblance of legitimation. Conversely it is plausible to suggest that a cult on the memory of Heroes and Ancestors could sustain a 'counter-memory' of the time before the Roman conquest and provide a focal point for the identity of native Britons into the Roman period (this identity in turn being subject to re-negotiation). In this respect the so-called Romano-Celtic temples would act to maintain this 'counter-memory' without there being any necessary continuity of religious function from the Iron Age to the Roman period. Coda: 51 Alban versus the NornUll'ls On the theme of domination and resistance the Norman Conquest provides an interesting comparison to the Roman conquest. In a similar fashion to the Romans, the Norman conquerors legitimated their rule as the inheritors of the mantle of the penultimate king of the native dynasty (Edward the Confessor equals Cunobelinus) rather than the last king (Harold equals Caratacus), who opposed their invasion. It is interesting that in this respect the cult of a sanctified King Edward was fostered at Westminster Abbey and was eventually the focus of the royal dynasties that had superseded his own. On the note of resistance and the negotiation of power between ruler and ruled in Britain, the appropriation of the memory of Edward the Confessor by the Normans was opposed by a native 'counter-memory' in which putative protagonists of the Saxon resistance such as the unfortunate Earl Waltheof (who became implicated in a plot by Norman nobles against William, and achieved the dubious distinction of being the only person officially executed, under Saxon law, in the reign of William 1) became objects of devotion after their death with miracles claimed at their tombs (Williams 1995:64). The last act concerns the relics of the Roman St Alban. According to tradition he was decapitated on the hill adjacent to the Folly Lane enclosure (with its own strange decapitated head ensemble) raising Whatever Happened to the Heroes? 95 suspicions that the legend of St Alban is related to that of the 'Hero on the Hill' at Folly Lane, and that the legend of St Alban and the legend of the Hero may have fused (Denison 1994; Niblett pers. comm.). In a last ironic twist the relics of St Alban were transported to Denmark, at the time of the Norman Conquest, by the fugitive abbot of St Alban's, and used as a talisman of resistance against William the Conqueror by the abortive liberation movement led by king 'Cnut the Holy' of Denmark (Williams 1995:47-48). School of Archaeological Studies, Leicester University Notes [I) The very notion of a broken gift seems quite perverse. Animal sacrifice is no exception. Here the gods were offered food - notoriously the more unsavoury portions, according to Hesiod - on the same terms as human beings. A preparation of a meal usually entails the prior destruction, dismemberment and cooking of the animal/plant. If this destruction, dismemberment and cooking is not regarded as 'mystical' in a secular context it difficult to see why it should be so in a sacred context. [2] A brief note on terminology: I hold that there is no ritual/utilitarian, irrational/rational distinction between a ritual pit and a well. A well is as 'ritual' an object as it is possible to imagine, as exemplified by the 'irrational' deposition of coins in them, up to the present day, by those unaware of archaeological methodologies which regard them, in the first instance, as merely functional utilitarian water-obtaining devices. 'Wishing' and 'holy' wells, in places such as the Peak district are the still the focus of annual (Christianised) rituals (Porter 1984:44, 144). Utilitarian practice does not preclude or have necessary priority over ritual practice. A cknowledgel11ents 1 would like to thank all those who commented on my paper when it was first delivered at TRAC 97, especially Rosalind Niblett, with whom I continued the discussion in the bar My inadequate knowledge of the classical world has been supplemented by Samantha Burke, to whom I express my thanks. I also wish to thank Mr Bartlett for his thoughts and for the tOllr of his delightful museum at Harlow. I hope have benefited from the advice of my anonymous referee and from Rob Witcher whose suggestions I have tried to take on board. I am also indebted to Kirsty Shipton for correcting my Latin and Greek. All mistakes, however, remain my own. Bib/iogmphy Ahern, E. 1973. The Cult of the Dead in a Chinese Village. Stanford: Stanford University Press. Bailey, C 1935. Religion in Virgil. Oxford: Oxford University Press. Bartleu, R. 1988. The Harlow Celtic Temple, Current Archaeology, I J 2: 163-66. Bedwin, O. 1981. Excavations at Lancing Down, West Sussex 1980. Sussex Archaeological Collections, 119: 37-56. Biddle, M. 1975. Excavations at Winchester, 1971: tenth and final interim report, Antiquaries Journal, 5596-126,295-337. Black, E. W. 1986. Romano-British Burial Customs and Religious Beliefs in South East England, The Archaeological Journal, 142:201-239. Braithwaite, G. 1984. Romano-British face pots and head pots, Britannia, 15:99-131. Branigan, K. 1973. Town and Country: The Arch.aeology of Verulamiilln and the Ronum Chilterns. Bourne End, Buckinghamshire: Spur Books. Burnham, B. C. and Wacher, 1. 1990. The 'Snwll Towns' of Roman Britain. London: Batsford. Burstow, G. P. and Holleyman, G. A. 1957. Excavations at Muntham Court, Findon Sussex, The