One for the Road? Providing Food and Drink for the Final Journey

ONE FOR THE ROAD? PROVIDING FOOD AND DRINK FOR THE FINAL JOURNEY EDWARD BIDDULPH How do archaeologists analyse pottery from a Romano-British cremation cemetery? Analysis is typically inquiry-led. How has vessel choice altered through time? How many pots per grave were deposited? Is their any correlation between the number of pots per grave and the types or sources of pots present? What differences are there between burial and domestic assemblages? Can the pottery reveal social differences within the burial population? All these are valid questions to ask of a dataset, and the answers provide understandings of funerary practice and the population of that cemetery itself. There must be a suspicion, however, that archaeologists are not gaining new understandings through analysis, but merely identifying trends in pre-understandings (cf. Hodder 1999, 49). For the ceramics, these 'pre-understandings' are about vessel function. Even before a burial is excavated, the excavator knows that it may well contain ceramic vessels of a reasonably limited type range. The excavator knows, too, that the function of these vessels concerns food and drink. Unfortunately, this 'pre-understanding' becomes embedded into classificatory and analytical processes, and prevents new understandings from forming. Let us bring this argument into focus by examining a common interpretation. In the report on the Roman cremation cemetery at Each End (Ash-next-Sandwich), Hicks discussed the treatment of accessory vessels. She suggested that many vessels would have contained food and drink to nourish the soul on the journey to the underworld (I 998, 115). The choice of vessels seems to reflect this. A typical cremation burial would be furnished with a liquid-serving vessel, normally a flagon or flask, a drinking vessel, such as a cup or a beaker, and a food-holding vessel, for example a platter, dish or bowl. The validity of this interpretation depends on two assumptions. First, that vessels of types found in burials were used in a dinner table setting. If we were invited to step back into the past and dine with a 101 EDWARD BIDDULPH Romano-British family, we would eat from a bowl, dish or platter, and pour wine from a flask or flagon into a cup or beaker. The second assumption is that the earthly functions of these vessels were retained upon their deposition within the burial pit. Just as in life, a cup, say, contained drink and was to be used in the afterlife as a drinking vessel. It is argued in this paper that the archaeology cannot always sustain these assumptions, and that alternative explanations for vessel use and treatment should be sought. This paper does not provide an exhaustive survey of alternatives, but instead suggests that the methods by which ceramic vessels are examined, inevitably determining our understanding of them, are inappropriate for funerary contexts. As the Each End cemetery provides the starting point of this paper, we will examine evidence primarily found in Kent. THE EVIDENCE Vessel combinations To equate to the 'dinner table setting', accessory vessels must conform to particular combinations (Biddulph 1996, 15). We should be able to remove the vessels from the ground, place them onto a dinner table and, once food and drink are provided, enjoy a meal. The scale of the meal, naturally, depends on the numbers of vessels present in the burial. An optimum range of vessels includes a flagon, a dish, and a cup. There may be variations as long as the broad functions remain the same. Thus, a flagon can be replaced with a flask, a dish with a bowl, and a cup with a beaker. A full meal, involving both eating and drinking, need not be represented. A burial may contain a flagon and a cup, or a dish only. In such cases, the soul can partake in certain elements of a meal. The soul can also drink if the burial contains a cup or beaker, or any number of both. Seven out of fourteen burials from the Each End cemetery fit part or all of the dinnerservice suite, including S.23, containing a flask and samian cup, and burial S.24, yielding a beaker, flagon, cup and dish (Savage 1998, 139-47). Many burials from other sites in Kent contained accessory vessels that also fit the model. These include a burial from Barming, in which a flagon, beaker and samian dish were deposited (Detsicas 1980, 396), two burials from Minster (Perkins I 985, 54-6), and most burials from Ospringe (Whiting et al 1931 ). Other vessel combinations are, by contrast, difficult to reconcile. From Wincheap in Canterbury, a burial produced two flagons and a beaker (Macpherson-Grant 1980, 102 ONE FOR THE ROAD? FOOD AND DRINK FOR THE FINAL JOURNEY 292-3). While the soul is able to drink, we may wonder why two flagons were deposited. Perhaps the journey to the underworld was particularly long. requiring two flagon's worth of refills, or that the drink was for sharing, either on the journey or at the destination, in which case, the absence of a second drinking vessel is curious. Burials that contain liquid-serving vessels only or liquid-serving vessels in combination with food vessels are more problematic. Such burials include eleven burials from Cranmer House, Canterbury (Pollard 1987, 285-295), and burials S.19 and S.22 from Each End, both containing a flask and samian dish (Savage 1998, 138-47). A beaker was deposited in the latter, but contained cremated bone, which alters its function from drinking vessel to urn. If these burials are transferred to the dinner table, the contents of the flagons or flasks could not be consumed without the presence of drinking vessels. However, perhaps we should not be overly concerned with this; in the spiritual world, the absence of tools to perform a function may not have mattered. It is also feasible that small flagons and flasks functioned as drinking vessels. Jars were placed in a number of burials as accessory vessels. Examples include burials S.13-15 at Each End (ibid.), burial groups 1, 25 and 44 at Cranmer House (Pollard 1987, 285-295), and burial groups 91, 100 and l 06 from Ospringe (Whiting et al 1931, 28-38). The jar functioned as a cooking vessel, as well as, say, a food-storage vessel, and may not always sit quite so easily within the dinner table setting. While the soul may be expected to pour liquid into cups, or even spoon food out from the jar, it was surely not expected to prepare its meal! The absence of mortaria from burial assemblages suggests that vessels are more likely to represent food consumption than food preparation, and we must assume that the jar found a regular place on the dinner table or concede that the earthly function of the jar was irrelevant to the funerary process. Function change Occasionally, accessory vessels were deposited or laid out in ways that suggest that the primary functions of those vessels altered (Biddulph 1996, 17). This is evidenced by the inversion of vessels and use as covers or lids. These are clearly deliberate actions, forming part of the ritual of burial. While we cannot know what such actions meant to the mourners carrying them out, it is likely that placing a vessel upside-down or covering the mouth of the urn added significance to the process of deposition. From the Cranmer House site in Canterbury, burial 28 included an inverted dish that covered the mouth of the urn (Pollard 1987, 291 ). The function of the vessel 103 EDWARD BIDDULPH changed from a food vessel to a lid; physically, it could not have borne food on its inversion. How may we interpret the choice and placing of this dish? Assuming that accessory vessels were symbolic of food offerings and did not necessarily contain them, then this interpretation is not affected. The dish could be placed upside-down, and the soul take nourishment from it. The inversion might have enhanced the role of the dish; what could be more natural than to offer food directly to the deceased by placing it over the urn? The physical contact between vessels ensured that the soul could take nourishment. Just one other accessory vessel, a beaker, was included in burial 28. Although small enough to have been placed within the urn, thus putting its contents in contact with the bones, it was placed outside the urn instead. To strengthen an interpretation, we have to assume that the treatment of vessels was consistent. Therefore, if the food was given directly to the deceased of burial 28, it follows that drink should also have been. An alternative interpretation might place significance on the covering of the urn, rather than the function of the dish. Wide and shallow vessels, chosen to cover the urns in burials I 02 and I 55 at Ospringe (Whiting et al 1931, 36; 66 ), served a secondary purpose, perhaps preventing loose soil from contaminating the bone, or providing a barrier against spiritual harm. That the use of a lid remains a relatively uncommon practice, spiritual, not mundane, motives appear to be behind it. Inverted vessels did not always provide cover. Again from Ospringe, the excavation of burial 63 revealed a samian bowl resting on its side against a jar (Whiting et al 1931, 16), while a burial from Cooling produced an urn that stood on top of an inverted bowl (Thornhill and Payne 1980, 381-2). Like the dish above, these vessels could not have held food. Deliberate mutilation Evidence for the deliberate mutilation of accessory vessels includes perforated bases or walls, broken handles or rims and entirely smashed vessels. Vessels with deliberate damage should not be confused with seconds. Seconds are accidentally misfired, but otherwise usable, products whose inclusion in funerary assemblages is well attested (cf. Tuffreau-Libre 2000, 54). Burial 137 from Ospringe, for example, produced a cup with a piece of its rim removed (Whiting et al 1931, 57 ). Archaeologists often say that such pottery has been 'killed' ( e.g. Going 1988, 22-3 ). A curious aspect of this practice is that it was infrequent and did not make up the norm. None of the 104 ONE FOR THE ROAD? FOOD AND DRINK FOR THE FINAL JOURNEY burials from Each End or Cranmer House includes 'killed' vessels, and less than ten per cent of Ospringe burials yielded mutilated pottery. Yet the practice was widespread, with examples found in such disparate areas as Winchester (Biddle 1967, 246), Great Dun mow in Essex (Wickenden 1988, 12-21 ), Chichester (Down 1971, 110), and York (Wenham 1968, 27). If the practice was widespread, then so was the meaning behind it and, rather than being a localised variation of burial practice, the ritual was the demonstration of, or originated from, a universally held belief. This belief is, of course, unknowable, but a popular interpretation is inherent in the label that we give to the evidence. In order to accompany the deceased into the underworld, accessory vessels must also pass into the realm of the dead by being 'killed'. Perhaps the vessels cannot serve their functions until this is done. Insuring against grave-robbing by making pottery unattractive to thieves, or providing the living with keepsakes, perhaps to bring good fortune, have been suggested as alternative interpretations (e.g. Going 1988, 22-3; Wenham 1968, 27). This physical action was applied inconsistently. A whole beaker was deposited in addition to the 'killed' cup in burial 137 from Ospringe. Even within cemeteries displaying a good range of evidence, such as Great Dun mow in Essex, burials contain a mixture of 'killed' and 'non-killed' vessels. This apparent mixture is not so curious, however, if we consider the possibility that vessels were 'killed' in archaeologically invisible ways. We may imagine a priest performing a ceremony allowing vessels of the living to enter the world of the dead. But, in whatever ways vessels were 'killed', the effect of deliberate physical or spiritual breakage is worth emphasising: the destruction of earthly utilitarian function. A flagon with a perforated base cannot hold liquid; we could drink from a fractured cup only with difficulty. A platter that has been made taboo for all outside the burial pit cannot be used in life. Since the idea that vessels carried food offerings depends on the vessels retaining both function and form in the burial pit, it is reasonable to suggest that if the primary function of a vessel could no longer be carried out in life, the vessel could not function in death. THE INTERPRETATIVE PROCESS We have seen how the evidence may not always be consistent with a single interpretation. Put simply, the problem is not so much that ceramic classification reinforces functional assumptions, but that 105 EDWARD BIDDULPH interpreters are failing to recognise that it does so. In the desire to reach an understanding of the broad picture - in this case, the functional character of a ceramic assemblage - we lose the significance of the 'pixels' on which that picture is based. And, in trying to understand the population, we risk failing to understand the individual. So, when we classify vessels into broad functional categories - jars, beakers, flagons and so on - we are creating interpretative straitjackets. The homogeneity of resulting groups allows for little recognition of variation within them (Shanks and Hodder 1995, 9). We fail to recognise when our evidence poorly fits the interpretation, and so we fail to countenance alternative interpretations. This is not to say that functional assumptions are wrong. Food offerings have been found in burials (Dobney 2001, 42). Besides, we have many ancient sources to furnish us with clues about how people lived. Pompeii and Herculaneum, for example, with their wonderfully preserved wall paintings and shop fronts, are excellent places to start. There we gain confirmation that flagons held wine and platters held food, and see that a host of other recognisable vessels were connected, unsurprisingly, with eating and drinking. We should be cautious when using this considerable wealth of information. The inversion and destruction of vessels are visible reminders that function and meanings could change with the burial act. If proof were needed that the archaeologies of the dead and the living do not necessarily share meaning and form, then this is it. Wall paintings and the like, while revealing much about how vessels were used in life, reveal very little directly about their use in death. Inevitably, many of our ideas of the past are constructed on the experience of the present. In the modern world, there are cups, jugs and plates used primarily for holding drink and food. We know this statement is true, because we have all experienced eating or drinking using these vessels. In the ancient world, there are vessels shaped like our cups, jugs and plates. It is often with reference to our experiences that we call them cups, jugs and plates, and give them the same functions as our similar looking modern vessels. This helps us to construct an understanding of the past, because we cannot experience ancient life first hand, nor can we ask a Roman or Briton to relate their experiences. This means that the modern personal perspective is inextricably linked to the interpretation of ancient material evidence. And, inevitably, differences result. The small samian vessel Drag. 27 is a ( drinking) cup for British archaeologists (e.g. Webster 1996, 38), but a sauce-dish for continental archaeologists (e.g. Schucaoy 2000, 123). We also know that from experience cups, jugs and plates are used in contexts other than the kitchen or dinner table. Cups serve as 106 ONE FOR THE ROAD? FOOD AND DRJNK FOR THE FINAL JOURNEY candleholders and receptacles for beads or buttons, plates serve as plant drip trays and artists' palettes, while jugs serve as plant pots, watering cans, and bathroom accessories. The list of alternative functions is endless. Understanding of the past is achieved through the transfer of modern experiences to past contexts (Hodder 1999, 45). And, if we are prepared to transfer generic experiences - plates for eating and flagons for liquid - we should also be prepared to transfer less widespread and more personal practices. That cremation vessels held sustenance for the journey to the afterlife is a popular and plausible interpretation. It has already been shown that burial evidence cannot always sustain this interpretation. In addition to vessel inversion and destruction, this evidence includes unburnt vessels, representing pottery not placed on the funerary pyre prior to burial, and is consistent with the view that unburned goods placed within the burial pit were deposited too late to accompany the spirit to the afterlife. A purpose of cremation was to release the spirit from the dead body. It was at this point that the spirit began its journey to the other world. It was necessary therefore, for any accompanying material, such as food and drink, to be placed alongside the body on the pyre, as all objects were transported to the other world by way of the rising smoke (Graslund 1994, 20). Interpretation of burial evidence tends to place too much emphasis on the vessels themselves, and not enough on their contents, if any. The reason for this is obvious: the evidence almost exclusively consists of the vessels, and rarely of what filled them. While it is a reasonably safe assumption that any substances that the vessels contained were likely to have been organic, those contents need not be connected with food and drink. For example, flasks might have used to pour ointments over the body before cremation, serving a purpose normally assigned to glass vessels; perhaps ceramic flasks were more affordable than glass bottles, and reflect the lack of status and wealth - the lives, in fact, of ordinary people. Flagons may have contained infusions not necessarily used as sustenance. Beakers may have held perfumes, such as frankincense and myrrh, while jars and dishes may have held plant material, such as poppies (symbolic of sleep and death), dates (symbols of reincarnation) and evergreens, symbolising eternal life (Kreuz 2000, 50). When function changes, a cup, for example, shares little more than shape with another cup. It may acquire intimate associations with its user that preclude anyone else from using it or understanding its meaning (cf. Miller 1987, 126). Vessels from cremation burials occasionally carry evidence that suggests alternative meanings beyond the mundane. A samian bowl in burial 156 from Ospringe was 107 EDWARD BIDDULPH inscribed 'LVCIVS LVCIANVS VLI DIANTVS VICTOR VICTORICVS VICTORINA VAS COMMVNIS', translated as the common dish of Lucius, Lucianus, Julius, Diantus, Victor, Victoricus and Victorina (RIB 2501.307). This inscription is unlike any other name inscription from the cemetery, as it provides proof of a relationship between the named individuals and the bowl in addition to the statement of ownership, although the nature of the relationship can only be guessed at. In life, the vessel may have been part owned and used by the named individuals. Perhaps one of them is buried in pit 156, and the bowl is symbolic of love, friendship, and eternal -togetherness. Possibly the vessel was bought specifically to be included in the burial as a communal offering. Whatever the vessel symbolises, it has acquired personal meaning in addition to its mundane function. While function is invariably physical, meaning is an abstract construction. It is reasonable to suggest, then, that vessels do not need to be inscribed to be given special meanings, and other vessels from Ospringe and every other cemetery may have carried meanings that can never be known. So, to give accessory vessels a single interpretation based on modern generic experiences is simplistic, and ignores the possibility of one or many different sorts of meanings and functions that may have been accorded to them. But if the choice of vessels reflects a multiplicity of motivations, often personal, why do many cemeteries display a narrow range of vessel types, to the extent of that range being standard? At Each End 'most of the multiple-vessel interments contained typical variations on the standard jar/flagon/dish/beaker ceramic suites which usually furnished cremations of the period' (Hicks 1998, 134). In claiming that there is a standard suite of vessels, interpreters fall into a trap. The standard suite is an inevitable result of classification by which means vessels are placed in pre-existing and pre-defined categories. This approach is eminently practical. The cemetery at Springhead, near Gravesend, produced over 600 vessels (Glass 1999, 207). It is clearly nonsense to give each vessel an indi victual classification, and would result in an unwieldy typology, masses of paperwork, and little hope of drawing from the process generalised statements about the past. By and large, vessels sorted into functional categories are separated on shape criteria. These are entirely subjective (this rim looks like a rim from a pot which has been called a beaker); although whole vessels can benefit from pseudo-objectivity- a beaker is small enough to be held in the hand and whose height is greater than its rim diameter. (Millett ( 1979, 37) concluded that this method of classification 'can be applied with reasonable consistency', but surely only if carried out by a single indi victual. There can otherwise be no 108 ONE FOR THE ROAD? FOOD AND DRINK FOR THE FINAL JOURNEY guarantee of repeatable results.) The danger is that similarities of shape are assumed to equal similarities of function. Consequently, there is no perceived difference in the meaning of burial vessels, nor motivation for their placement within the burial pit. And, as the evidence presented above has suggested, similarly shaped burial vessels can serve quite different functions or have individual significance. There is a solution. Acknowledge that functional assumptions are merely assumptions. By all means, sort the 600 Springhead vessels into jars, bowls, dishes, beakers, and so on. Use the resulting dataset to compare one burial assemblage to another (e.g. Millett I 993) and see how it differs from a domestic assemblage (e.g. Schucany 2000). Recognise that these categories are nothing more than convenient labels. To gain an understanding of the significance of the vessels that constitute the assemblage, we must return to each burial pit. At this point - the closest point to the original funerary process - we can impose meaning on the vessels. So, an inverted wide and shallow vessel placed over the mouth of the urn is labelled a dish, but in functional terms it is a lid, because the vessel no longer functions as a dish. At once, the food connotations are disregarded. Similarly, a beaker placed inside an urn should now carry no explicit drink-related meaning, but a purely conceptual meaning based on its placement. Perhaps it should be classified as an ·insider'. An inscribed bowl falls within the 'personal relationship' category, not bowl/eating category. So, vessels can be classified in terms of their position or individual treatment within the burial pit. In re-classifying the material, we de-emphasise the food element and re-focus on the rite itself. We may then be able to gain better understanding of vessel treatment and, ultimately, of motivation for vessel choice. CONCLUSION We have seen how the seemingly standard range of evidence from cremation burials does not always fit a single, albeit popular, interpretation. Two burials containing the same numbers and types of vessels may be linked only by physical coincidence, but the reasons for the choice of vessels and their contents may differ substantially. A single burial, too, may contain a palimpsest of meanings; food and drink may play just a small role within a suite of beliefs that dictate vessel choice. Material absences, for example vessel contents, are but one constraint to a fuller understanding of the past. Modern perception, although inevitable, is another, and is compounded by the 109 EDWARD BIDDULPH way that we classify material. Conventional typologies are perhaps an inappropriate way to classify the archaeology of the dead, and risk losing the connection between the vessels and the rite. We need to invent a typology of burial that incorporates inversions, vessel placement, mutilations and inscriptions into the classification process, so that we do not lose these 'conceptual' data from the analytical process. ACKNOWLEDGEMENTS The author is indebted to Kim Biddulph and Dr Neil Faulkner for reading draft texts and providing invaluable comments and advice, and to Dr Andrew Gardner for undertaking some of the 'leg-work'. BIBLIOGRAPHY Biddle, M., 1967, 'Two Flavian burials from Grange Road, Winchester', The Antiquaries Journal, xlvii, 224-50. Biddulph, E., 1996, 'The Vessels from Romano-British Cremation Burials: a reassessment of classification and interpretation', unpublished MA dissertation, University College London. Detsicas, A. P., 1980, 'A grave group from Barming', Archaeologia Cantiana, XCVI, 396. Dobney, K., 2001, 'A place at the table: the role of vertebrate zooarchaeology within a Roman research agenda for Britain', in S. James and M. Millett (eds), Britons and Romans: advancing an archaeological agenda, Council for British Archaeology Research Report 125, York, 36-45. Down, A., 1971, 'The Roman cemetery at St Pancras', in Down, A. and Rule, M., Chichester Excavations, vol. I, Chichester Civic Society Excavation Committee, Chichester. Glass, H. J ., 1999, 'Archaeology of the Channel Tunnel Rail Link', Archaeologia Cantiana, CXIX, 189-220. Going, C. J., 1988, 'Ritual', in Wickenden, 22-3. Graslund, B., 1994, 'Prehistoric soul beliefs in Northern Europe', Proceedings of the Prehistoric Society, Ix, 15-26. Hicks, A. J., 1998, 'Excavations at Each End, Ash, 1992', Archaeologia Cantiana, cxvm, 91-172. Hodder, I., 1999, The Archaeological Process: an introduction, Batsford, London. Kreuz, A., 2000, 'Functional and conceptual archaeobotanical data from Roman cremations', in Pearce et al., 45-51. 110 ONE FOR THE ROAD? FOOD AND DRINK FOR THE FINAL JOURNEY Macpherson-Grant, N. C., 1980, 'The pottery', in P. Bennett, N. C. Macpherson- Grant and P. Blockley, 'Four minor sites excavated by the Canterbury Archaeological Trust, 1978-1979', Archaeologia Cantiana, XCVI, 267-304. Miller, D., 1987, Material Culture and Mass Consumption, Blackwell, Oxford. Millett, M., 1979, 'An approach to the functional interpretation of pottery', in M. Millett (ed.), Pottery and the Archaeologist, Institute of Archaeology Occasional Publication, 4, London, 35-48. Millett, M., I 993, 'A cemetery in an age of transition: King Harry Lane reconsidered', in Struck, M. (ed.), Romerzeitliche Grttber als Quell en zu Religion, Bevolkerungsstruktur und Sozialgeschichte, Archaologische Schriften des Instituts fiir Vor- und Frllhgeschichte der Johannes Gutenberg- Universitat Mainz, 3, Mainz, 255-82. Pearce, J., Millett, M. and Struck, M., 2000, Burial, Society and Context in the Roman World, Oxbow, Oxford. Perkins, D. R. J., 1985, 'The Monkton gas pipeline', Archaeologia Cantiana, CII, 43-70. Pollard, R. J., 1987, 'The pottery: (I) The cremation burials', in S. S. Frere, P. Bennett, J. Rady and S. Stow, Canterbury Excavations: intra- and extra-mural sites, 1949-55 and 1980-84, The Archaeology of Canterbury, viii, Maidstone, 285-95. The Roman Inscriptions of Britain, Vol. II, fasc. 7, 'Graffiti on samian ware', edited by S. S. Frere and R. S. 0. Tomlin, Alan Sutton Publishing, Stroud, 1995. Savage, A., 1998, 'The Roman pottery', in Hicks 1998, 132-50. Schucany, C., 2000, 'An elite funerary enclosure in the centre of the villa of Biberist-Spitalhof', in Pearce et al., I I 8-24. Shanks, M. and Hodder, I., 1995, 'Processual, postprocessual and interpretive archaeologies', in I. Hodder, M. Shanks, A. Alexandri, V. Buchli, J. Carman, J. Last and G. Lucas (eds), Interpreting Archaeology: finding meaning in the past, Routledge, London, 3-29. Thornhill, P. and Payne, P., 1980, 'Some sites in North Kent', Archaeologia Cantiana, XCVI, 378-80. Tuffreau-Libre, M., 2000, 'Pottery assemblages in Gallo-Roman cemeteries', in Pearce et al. 2000, 52-60. Webster, P., 1996, Roman Samian Pottery in Britain, Council for British Archaeology, York. Wenham, L. P., 1968, The Romano-British Cemetery at Trentholme Drive, York, HMSO, London. Whiting, W., Hawley, W. and May, T., 1931, Excavation of the Roman Cemetery at Ospringe, Kent, Reports of the Research Committee of the Society of Antiquaries of London, vii, Oxford. Wickenden, N. P., 1988, Excavations at Great Dunmow, Essex: a RomanoBritish small town in the Trinovantian civitas, East Anglian Archaeology Report no. 41. 111