Gladiators at Roman Colchester: Re-Interpreting the Colchester Vase

Abstract

The locally made colour-coated ware vessel known as the Colchester Vase is argued to be a commissioned piece recording a performance in the town. The inscription on the vessel, cut pre-firing, names individual arena performers depicted en barbotine. One name, Memnon, is argued to be a ‘stage name’ taken from a protagonist in the Trojan war. The connection of another combatant, Valentinus, to the 30th legion is re-considered as evidence for gladiators linked to the Roman army. The Vase's final use was as a cremation urn. Osteological and isotopic analysis reveals the cremated remains to be those of a non-local male of 40+ years; unlikely to be one of the performers, he may nevertheless have been closely connected to the event.

INTRODUCTION¹

The Colchester Vase is a large colour-coated ware beaker made c. a.d. 160–200 in the kilns to the west of the town, decorated en barbotine with arena scenes and inscribed with the names of the protagonists (fig. 1). The companion paper to this article suggests that it was a product of the workshop to which the named potter Acceptus iii belonged, active in Colchester in c. a.d. 160–200.² It was used as the cremation urn in a burial group of four vessels found in 1853 in the town's western cemetery area.³ The burial group's discovery, the late nineteenth- and early twentieth-century interpretations of the scenes and inscription, a detailed description of the scenes and of the vessel's method of manufacture and stylistic affinities can be found in the companion paper.⁴ Re-examination of the lettering, it will be argued, identifies the inscription as being cut before firing rather than afterwards, as current orthodoxy has it. Setting the inscription and images in the wider body of evidence for games enriches the insights which can be derived from the names and associated information into arena performers in the north-west provinces. Undertaken as part of Colchester Museums’ Decoding the Dead project, osteological and strontium isotope analysis of the well-preserved cremated bones within the vessel enable them to be identified as the remains of a non-local male of 40+ years. Drawing on our study and its companion paper, the biography of the vessel and its implications for the understanding of the provision and celebration of gladiatorial games in Colchester are explored.

FIG. 1. The Colchester Vase burial group, including mortarium lid, dish and flagon, COLEM:PC.727–730. (Photo: D. Atfield; © Colchester Museums)

THE INSCRIPTION

The short inscription on the Colchester Vase naming arena performers has attracted attention since its discovery, with early commentators and museum visitors puzzling over its interpretation.⁵ Its essentials were set out by Hübner in his 1873 corpus of inscriptions from Britain; the RIB edition published in 1993 corrected some misconceptions which had developed in the intervening period.⁶ However the following re-evaluation, enabled both by autopsy and by new higher resolution images of the inscription, challenges current orthodoxy concerning the text's creation. The evidence of lettering is examined first before re-assessing the names of the performers and their associated attributes.⁷

LETTERING

The inscription is incised a little below the rim of the vessel, being placed where the upper frame bounding the decoration is interrupted. It comprises four captions divided into two pairs of names, one pair (SECVNDVS MARIO) placed above the whip-carrying figure and associated with the animal combat, the other (MEMNON SAC VIIII VALENTINV LEGIONIS XXX) labelling the secutor and retiarius (fig. 2). It is confidently cut in fine strokes by a stylus or similar finely pointed instrument in letters a few millimetres high, the same hand being responsible for the whole text. Two names, those of Secundus and Mario, are incised close to the rim in smaller letters (mostly 4–6.5 mm high), while the others, those of Memnon and Valentinus, are incised at a greater distance from the rim and in larger letters (mostly 7–8 mm high). Letters are executed in fluid rustic capitals, generally incised in two or more straight or lightly curving strokes. The superimposition of strokes observed under magnification allows their general sequence of execution to be established, illustrated for example by the unbarred A, S, E and I. Typically a short vertical or diagonal stroke is cut first, followed by further horizontal or diagonal strokes, and the letter is finished with serifs.

FIG. 2. The inscription below the rim of the Colchester Vase. (Photo: D. Atfield; © Colchester Museums)

The writer's confidence is especially expressed in the more mannered lettering in the captions for Memnon and Valentinus, facilitated by greater distance from the potential obstruction posed by the vessel rim. It manifests in flourishes like the lightly curved and tapering serifs on the diagonal and vertical strokes of E, I, N, X and M, or in the curling leftward hooks I in Valentinus or L in legionis (figs 3–4). The flamboyant upward diagonal stroke from the S of legionis and the upwards curl of the bottom serif of the R of Mario, made without lifting the stylus, also exemplify the writer's élan (fig. 5). The diagonals of contrasting width which render X illustrate the alternation between broader and narrower strokes of which the writer was capable, achieved perhaps by changing the angle of the stylus tip. Their contrast echoes, for example, the variation of thick and thin strokes in rustic capitals and finer cursive writing at Vindolanda.⁸ Some letter forms, for example the S and L of legionis, indicate a wider affinity with Old Roman Cursive writing.⁹

FIG. 3. The Colchester Vase inscription. Detail of the word VALENTINV. (Photo: D. Atfield; © Colchester Museums)

FIG. 4. The Colchester Vase inscription. Detail of the words LEGIONIS XXX. (Photo: D. Atfield; © Colchester Museums)

FIG. 5. The Colchester Vase inscription. Detail of the word MARIO. (Photo: D. Atfield; © Colchester Museums)

The Colchester text is classified by RIB with inscriptions incised on vessels after firing, but the calligraphic quality and fluidity of its execution is not otherwise paralleled in this category.¹⁰ Its finesse has in fact largely gone unremarked, despite Hübner's assessment of the evident expertise (‘litteris optimis’).¹¹ The identification by Roach Smith of the text's incision as being ‘posterior to the fabrication of the urn’ has been followed by RIB's editors and has gone unchallenged in other publications, whose authors may not have examined the vessel in person.¹² However, the absence of ragged edges on the letter strokes indicates that they were cut before firing, and the lettering is also reminiscent of other texts cut at the same stage, for example the kiln tallies from La Graufesenque and the cursive and capital texts incised on the leather-hard clay of unfired tiles.¹³ The suppleness of the Colchester Vase inscription also recalls similar lettering in other malleable media, for example on some lead curse tablets.¹⁴ The deciding evidence, however, for this being a pre-firing inscription is the superimposition of the strokes which form individual letters (as noted above). This could not conceivably be achieved after firing. The lettering was therefore incised before firing into the plastic medium of the drying clay, following the application of relief decoration and the colour-coat slip. The text was thus part of the vessel's conceptualisation from the start and relates directly to the scene portrayed, relieving doubts that it might be a later addition to a generic arena representation.¹⁵

TEXT AND COMMENTARY

SECVNDUS MARIO // MEMNON SAC VIIII // VALENTINV LEGIONIS XXX

Secundus Mario // Memnon s<A=e>c(utor) VIIII (pugnarum) // Valentinu(s) legionis XXX

‘Secundus. Mario. Memnon secutor, nine fights. Valentinus of the 30th legion, (tiro).’

The majority of the text comprises unabbreviated single names, the commonest form of text incised on ceramics in Britain.¹⁶ It is one of only two inscriptions from Britain that name arena combatants,¹⁷ making inscriptions naming performers from other provinces the most appropriate comparanda. Among the latter single names occur much more frequently than names in other formats.¹⁸ These single names in cognomen form raise the question as to whether they precisely identify the status of the performers, for example as slaves or peregrini,¹⁹ or simply supply the stage names behind which a real status can only be surmised.

The placing of Secundus and Mario above the head of a single figure is best explained by the restriction of the inscription to spaces on the vessel without an upper framing line (see above). Secundus is a commonplace name, but Mario is much rarer, not a Celtic name (pace RIB II) but ‘un nom latin rarissime’, attested primarily among current and former slaves on inscriptions from the central Mediterranean.²⁰

Like most of the (relatively few) captions labelling arena performers working with animals, Secundus and Mario are referenced by name only, without ancillary information. By contrast, the addition of type, combat history and familia affiliation for the two gladiators, Memnon and Valentinus, provides more material for discussion. This legend has usually been read as recording two different sorts of information, for Memnon his gladiatorial type and the number of fights and for Valentinus his ownership by the 30th legion, based at Xanten from a.d. 122 onwards. This interpretation is cautiously preferred in the discussion below, but the inconsistency it assumes in the format of the captions linked to the pair of gladiators should be noted.²¹ In what follows, therefore, we argue that different kinds of information are provided for the different fighters, though ambiguities remain.

Memnon is attested occasionally as a personal name in Latin inscriptions (mainly in Italy as well as on the Danube) and a little more frequently in Greek inscriptions of late Republican and imperial date from the eastern Mediterranean; its significance is considered further below.²² The abbreviation SAC is usually corrected and expanded to secutor, describing his fighting style.²³ Given the image of the secutor above which the inscription is incised, this is the most likely expansion, though an alternative reading of an abbreviation for sca(eva), ‘left-hander’, correcting SAC to SCA, also has some merit. VIIII must refer to the number of bouts fought during his career to date (placing him perhaps in mid-career), a single numeral without further qualification having this significance from the late first century a.d. onwards.²⁴

The interest of the name of the secutor, Memnon, has not been previously recognised. Given its scarcity, the name Memnon seems unlikely to be derived from local naming practice or to be a chance adoption from the Greek name repertoire. The mythical bellicose king of the Ethiopians, with a martial reputation second only to Achilles and slain by the latter at Troy, supplies a much more likely heroic exemplar for a stage name.²⁵ Stage names for gladiators based on heroes of epic linked to Troy, such as Patroclus, Achilles, Ajax and Idomeneus are well documented in the eastern Mediterranean; their occurrence in the north-west provinces is occasional, for example Hector and Ajax respectively on a tombstone from the Jura and an Augsburg mosaic.²⁶

As likely givers of stage names it is reasonable to assume a knowledge of epics of the Trojan cycle on the part of the elite owners or administrators of gladiatorial troupes, for example the equestrian procurators responsible for the emperors’ familiae in the provinces.²⁷ As for the specific figure of Memnon, dedications scratched by Egypt's equestrian prefects onto the Colossi associated with the hero at Thebes (among which at least one, Haterius Nepos, had served in Britain), witnessing that they had heard them ‘sing’, illustrates elite interest in the traditions attached to Priam's nephew.²⁸ The recently excavated mosaic from Rutland provides a vivid example of the familiarity of Roman elites in Britain with the stories of Troy.²⁹ A popular resonance of such ‘stage’ names in Rome's north-west provinces can also be argued. For example, among the mythological scenes depicted on mid-imperial Rhône valley ceramic medallions, one renders the death of Hector at Troy with legends identifying participants.³⁰

Famed for his beauty and valour, Memnon's blackness was repeatedly referenced in Roman period literary tradition along with his ‘Ethiopian’ origins; perhaps the appearance of this performer, especially his skin colour, may have influenced the giving of the stage name.³¹ Gladiatorial stage names sometimes drew on a performer's physical attributes, emphasising good looks, for example Callimorphus, as well as specific individual characteristics.³² For instance, the gladiators Xanthus (‘blondie’) and Anthrax (‘coal-black’) – the latter a term sometimes used in Greco-Roman literature as a descriptor of black people's skin – named on Rhône valley ceramic depictions of combat contemporary with the Colchester Vase provide regional instances of this naming principle.³³ The only other documented use of Memnon as a performer's name, for an Egyptian charioteer in late antique Carthage, clearly takes its cue from the skin colour of the rider celebrated in verse.³⁴ The latter case suggests that the choice of name for the Colchester performer might point only in the most general terms towards ‘Memnon's’ personal or familial connections to northern Africa, rather than towards any specific geographical origin, especially given the mutability in the placing of ‘Aethiopia’ in the Greco-Roman world-view.³⁵

Valentinu(s) as a cognomen is documented in six cases in Britain and very widely beyond.³⁶ From the lack of reference to previous fights he can be identified as a tiro, as first suggested by Herbert Blakiston, President of Trinity College, Oxford, in the early twentieth century.³⁷ This need not mean that the odds were stacked hopelessly against him: for example, at Trajan's triumphal games the tiro Marcus Antonius Exochus, trained in the imperial ludus at Alexandria, defeated an opponent of nine fights standing.³⁸ The reference to Legio XXX may indicate that Valentinus belonged to a familia of gladiators in the possession of this legion, which was garrisoned at Xanten from c. a.d. 122 into Late Antiquity.³⁹ This has prompted Jocelyn Toynbee and others subsequently to suggest that legions had their own troupes of gladiators and associated performers.⁴⁰ Some other evidence supporting this argument has seemingly been overlooked, the most significant being fragmentary third-century a.d. papyri implying the attachment of gladiators to the legionary garrison at Babylon, i.e. Old Cairo.⁴¹ Conjecturally the army might also have had some responsibility for the imperial familiae of gladiators for the provinces of Britain, the Germanies and elsewhere, attested on inscriptions which celebrate the careers of their procuratorial overseers.⁴² Yet there are reasons for being cautious over claiming a widespread military role in the training and supply of gladiators. Direct evidence remains very rare and gladiators were not a source of military recruits, except as a last resort (e.g. in a.d. 69, Tac., Hist. 2).⁴³ Instead, Valentinus’ affiliation to this particular legion might be a consequence of its being garrisoned in a region playing a nodal role in the procurement and supply of animal and human performers for the north-west provinces. The cities and garrisons of the lower Rhine possess famous inscriptions hinting at soldiers’ involvement in large-scale animal capture for the arena,⁴⁴ while the larger Rhineland cities dominate the limited corpus of epigraphic evidence from north-west Europe for arena performers, trainers and so on (see below).⁴⁵

THE CREMATED REMAINS by Emily Carroll

THE INDIVIDUAL

The cremated remains from the Colchester Vase were analysed as part of the Decoding the Dead project.⁴⁶ Overall, the cremation deposit was sufficiently well preserved to enable 51 per cent of the recovered burned bone to be identified to skeletal site (439.8 g of the 856.8 g in total). One individual was identified in the deposit. Demographic assessment suggested that their osteological sex is male and that they were aged over 40 years at time of death. The entire skeleton was well represented with the majority of identified bone (45 per cent, 199.2 g) derived from the skull. Full details are given in the Supplementary Material.

The individual had several vertebrae demonstrating extra bone growth or lipping around the vertebral body. These are referred to as osteophytes that usually form around joints affected by osteoarthritis, a condition predominantly associated with older individuals.⁴⁷ With regard to pathology, active periostitis was identified on the proximal femur, suggesting active inflammation of the bone's periosteum at the time of death. This lesion is demonstrative of non-specific infection, strain or trauma. In addition, six parietal fragments from the skull showed signs of porosity and pitting, associated with porotic hyperostosis which may be indicative of a nutritional deficiency.

THE CREMATION PROCESS

An interesting characteristic of the cremated bone recovered from the Colchester Vase is the level of fragmentation observed. Overall, the fragments were large, over half measuring more than 10 mm in size. A considerable proportion of the material thus has identifiable features preserved, which enabled sex and age assessment. Thompson and colleagues observed during their examinations of Romano-British cremation burials that a high level of bone fragmentation indicates that the burned bone was repeatedly stirred or mixed within the pyre.⁴⁸ The lesser fragmentation of the cremated bone from the Colchester Vase suggests that the pyre in this case was not well maintained and was left to burn down with minimal intervention. This is typical of urban cremations in Roman Britain, which may indicate that the practice of cremation became ‘industrialised’ as a result of the introduction of ustores (professional cremators), where pyres were not as well maintained when compared to rural practices.⁴⁹

Bone undergoes a progression of colour change when subject to extreme heat as a result of the combustion of bones’ organic and inorganic components.⁵⁰ While this progression varies slightly depending on temperature, time, and oxidisation, the sequence is always consistent.⁵¹ The majority of bone recovered from the Colchester Vase was predominantly white, with blue and grey pigments throughout the assemblage. This indicates high burning intensity but inconsistent firing conditions. McKinley has stated that this level of variation in burned bone colour is ‘normal’ and is a reflection of issues concerning oxygen supply, duration and the temperature of the fire.⁵² Coupled with the general lack of fragmentation and the relatively high status of the burial, this information suggests the potential use of professional cremators.

THE STRONTIUM ISOTOPE ANALYSIS by Joanna Moore, Geoff Nowell and Janet Montgomery

METHOD

Strontium isotope analysis of the cremated remains from the Colchester Vase was also undertaken as part of the Decoding the Dead project. Powdered bone samples were collected from the otic capsule and leached in 0.1M acetic acid following the standard protocol.⁵³ Strontium purification was achieved using column chemistry methods⁵⁴ at the Arthur Holmes Isotope Geology Laboratory (AHIGL), Durham University. Strontium isotope ratios were determined by Multi-Collector Inductively Coupled Plasma Mass Spectrometry (MS-ICP-MS) using a ThermoFisher Neptune MC-ICP-MS; the samples were introduced into this using an ESI PFA50 nebuliser and a micro-cyclonic spray-chamber. Reproducibility of international isotope reference material NBS987 was 0.710258 ± 0.000018 (2sd, n=10), and all data was normalised to an accepted value for NBS987 of 0.71024.⁵⁵

RESULTS AND INTERPRETATION

The sample obtained from the petrous bone of the Colchester Vase cremation (sample ID COL_18) has a strontium isotope ratio of 0.710842 ± 0.000011 (2 SE) and strontium concentration of 121.9 ppm, which are within the ranges expected for Britain.⁵⁶ Colchester is situated on the geology of the London Clay Group, a region of varied lithology comprised of sands, gravels and clays. These types of rocks are expected to produce strontium isotope ratios ranging between 0.709 and 0.7105 in the plants growing in these geological regions.⁵⁷ In the absence of a more detailed local strontium biosphere map, this range can be considered representative of the local environment. It is expected that people who source most of their food and drink from within this region would have strontium isotope ratios close to this range. As can be seen in fig. 6, the majority of the Colchester individuals analysed for the Decoding the Dead project plot within this range and closely match other Romano-British populations from regions with similar geology.⁵⁸ COL_18 exhibits a higher strontium isotope ratio than expected for local origins in Colchester; such values indicate possible origins in regions of Britain with older geology such as the south-west coast of England, parts of Wales, Cumbria and Scotland.

FIG. 6. Colchester cremation strontium isotope data alongside regional comparative data (Shaw et al. 2016). The horizontal dashed lines represent the bioavailable strontium isotope range for Essex (Evans et al. 2010). The analytical error for ⁸⁷Sr/⁸⁶Sr is within the symbol.

It is difficult to narrow down possible places of origin using a single isotope system, especially in Roman cities where populations are considered to have been highly cosmopolitan with high levels of mobility.⁵⁹ Colchester had been a major Iron Age centre, a legionary fortress and then the first provincial capital of Britain, and it would have been home to many different people and populations originating from across the Empire.⁶⁰ The town had early links with Legio XX, which recruited from Italy as well as other Mediterranean provinces.⁶¹ As such it is important to note that strontium isotope ratios consistent with Essex can also be found in other regions of the Roman Empire and on their own may not be entirely indicative of local origins. As well as a potential origin elsewhere in Britain, the higher strontium isotope ratios seen in individual COL_18 from the Colchester Vase could also suggest childhood origins outside Britain in other regions of the Roman Empire, such as northern Italy, northern Greece, Pannonia and Bavaria.⁶²

DISCUSSION

As well as being the site of the only known Roman circus in Britain, Colchester has produced many items referencing spectacle entertainment, some mass-produced memorabilia, others locally made: glass sports cups, lamps, relief-decorated pottery and moulds, including a mould-punch for a gladiator figure, a terracotta plaque, a wall painting showing a defeated hoplomachus(?) and a knife handle in the form of a victorious murmillo.⁶³ This diversity of visual evidence, supplemented by the evidence of investment in buildings which could accommodate games – one theatre is documented within the town, another in its environs at the Gosbecks sanctuary as well as the circus – speaks to the affinity of the Victricenses with Roman spectacle culture. It also speaks to their ability to understand the significance of the texts and actions represented on the Vase.⁶⁴

Reassessment of the Colchester Vase has major implications for interpreting the biography of the object and for informing the broader discussion of Roman spectacles taking place at Colchester, as it is best interpreted as commemorating a real event in the town. That it was a local product of the Colchester kilns, that the inscription was cut pre-firing and that the combatants are depicted in distinctive detail together suggest that the Vase was a commissioned piece and that the named individuals were real performers.⁶⁵ Precise observation of their arms and armour suggests a sufficient familiarity with Roman spectacle to notice and document the idiosyncrasies of the garb, arms and identity of participants, including the left-handedness of the secutor Memnon, as well as the distinctive dress of the animal fighters.⁶⁶ The exceptional quality of the barbotine decoration as well as the finesse of the writing is, moreover, indicative of a potter/potters meeting a commissioned brief as opposed to operating from a limited design repertoire.⁶⁷ It is conceivable that the individual/s who made the Vase were among the audience.

The staging of such spectacles in Colchester was likely enabled by its privileged position on long-distance communication routes. From his home at Xanten, the journey of Valentinus by land and by river and sea vessel to Colchester would have taken less than seven days under optimal conditions, eased by long-established trade networks linking his destination, a few miles inland on the Colne, with the Rhine and its tributaries.⁶⁸ These cross-channel links in the movement of goods and people between south-east Britain and the Rhineland are epitomised in the merchants’ dedications at the Rhine mouth shrines to the goddess Nehalennia.⁶⁹ Many of those traders originate in the riverside cities on the Rhine and Moselle which possess visual evidence in the towns proper and their hinterlands for munera and venationes.⁷⁰ This transmarine connection might have brought arena performers of all kinds to Colchester. The human mobility hinted at in the onomastic evidence discussed above, especially for Mario and Memnon, is not surprising in the context of the general mobility of gladiators attested in epigraphic and occasionally visual sources, now supplemented by isotopic evidence from human skeletal material.⁷¹ The pairings might even illustrate the orchestration of the diversity of performers for stage effect noted by Valerie Hope: ‘Perhaps pitting men of different nationalities and towns against each other added to the tensions of the arena's combats’.⁷² The same cross-channel connections might also have supplied animals for arena spectacle, above all the bears embodied in the creature abused by Secundus and Mario, as the likely most spectacular bestiae / ferae to appear in the British arena. Inscriptions indicate both the scale of wild animal procurement in the lower Rhine and military involvement in animal capture. Epigraphic testimony includes a second-century a.d. dedication to Diana recording the taking of 50 bears by one centurion, the vivarium or animal enclosure established by another in the late first or early second century a.d. (the only example documented outside Rome) and a later dedication to Silvanus by the bear catcher (ursarius) of the 30th legion at Xanten.⁷³ This regional specialisation may explain the zooarchaeological testimony for bears abused in venationes in Rome's northern arenas.⁷⁴

Adding name captions to arena and other spectacle representations has a long pedigree in Roman art.⁷⁵ Mosaics and paintings thus annotated survive occasionally in northern and western Europe.⁷⁶ Captioned portable objects from the same regions are more numerous, above all in serial productions like glass sports cups and more occasionally samian vessels.⁷⁷ These are, however, likely to be generic representations of spectacles: the Colchester Vase instead belongs among the small corpus of inscribed objects made to mark specific arena events. For example, thin-walled cups from the atelier of Gaius Valerius Verdullus in the middle Ebro valley celebrated gladiatorial and other spectacula at Calagurris (Calahorra), naming gladiators, their types, familia and fight outcome, as well as the editor (and vessel maker) for a specific occasion.⁷⁸ Representations of combats on the second- to third-century medallions for vessels in the Rhône valley give similar detail as well as combatants’ fight histories.⁷⁹ Other one-off mementoes include barbotine-decorated vessels from the fort at Langenhain, near Wiesbaden, and the Hees cemetery at Nijmegen, the former labelling gladiators and musicians, the latter a pair of combatants, some of these performers also with likely stage names.⁸⁰

Despite its affiliation with this captioning tradition, the Colchester Vase differs somewhat from other examples, especially in the anticipated interaction between viewer and object.⁸¹ The colour contrast between the slip and the fabric as well as the larger size of lettering on the Colchester Vase facilitated reading at a distance – other vessels noted above would have needed closer scrutiny to read the text, given their small lettering.⁸² Some would also have required rotation to appraise scenes and captions fully, whereas the scenes on the Vase with captions were (just) visible in their entirety in a single field of view. These characteristics potentially gave a more monumental impact to the Colchester vessel, enabling possible legibility from a few metres away, albeit dependent on lighting.

Whether deliberately or not, the Colchester vessel also differs in its presentation of narrative. On the Langenhain and Nijmegen beakers, for example, fights are in progress, neither inscription nor image supplying the outcome; on the Rhône and Ebro vessels, the texts report the outcome of the battle which continues in the associated scenes. On the Colchester vessel, the ad digitum gesture by Valentinus signals the end of combat, but the decision to spare or slaughter him, and thus his fate as victim or survivor of his first combat, is unreported. At the banquets whose voluptas it enhanced or at the funeral in which it was eventually interred, the Colchester Vase invited those who encountered it to relive the ‘collective effervescence’ of the spectators and the drama of the fight's dénouement.⁸³ The recall of Memnon might have prompted other reflections, given the likely sensitivity of some viewers to the literary pedigree of his name and to the resonance of a hero linked with one coast of Ocean winning in combat on its opposite shore.⁸⁴

As a commemorative piece of sports memorabilia, the original use of the Vase before its interment as a cremation vessel raises interesting questions, especially given its size, limited evidence for its handling and its findspot. Although munera originated in funeral contexts, it is unlikely, considering the late date of the vessel, that it was commissioned solely as a funerary urn depicting a munus that had publicly honoured the deceased.⁸⁵ Furthermore, the Vase also depicts an animal hunt (venatio), an activity that is not otherwise documented as an element of funerary ritual.⁸⁶ Vessels of this type are also only rarely documented in grave furnishing. There are only a very few graves from which barbotine vessels have been recovered from burials at Colchester and all were ancillary grave goods.⁸⁷ Indeed, the only other comparable decorated barbotine vessel from Britain, used as a cremation vessel, is a hunt-cup from Bedford Purlieus in the Nene Valley.⁸⁸ If not therefore commissioned as a cremation vessel, the Vase may instead have been destined for display, testament to the commissioner-owner's status.

The Colchester Vase displays little original use-wear, if any, and, as observed by Margaret Darling, it is questionable whether such exceptionally large beakers would have been of practical use in a convivial context. Empty it weighs over 1 kg; filled with a liquid it would be extremely heavy, cumbersome to pass around and difficult to drink from.⁸⁹ This lack of use adds weight to the Vase being a souvenir with which contact was limited to viewing and occasional turning, perhaps displayed in a setting of hospitality by its owner, prompting memory of the events depicted and perhaps discussion of the owner's connection with them. This mnemonic potential of the Vase within a hospitality setting may also have been to the fore in its final display and deposition within the grave at the funeral ceremony. While its furnishing does not strongly distinguish this grave from contemporary burials in the same cemetery, the Vase's very specific biography and its direct allusions to past events may have prompted memories among funeral participants of this individual's status.⁹⁰ As noted above, only in one other case from Britain is such a vessel used as a burial container, the slightly larger beaker from Bedford Purlieus showing venatores fighting a bear and a stag. In this case, evidence survives for the vessel's association with a monumental tomb – suggesting an occupant of high status.⁹¹

The Vase was undoubtedly an expensive product, produced by a ‘master potter’, and it is plausible that it was commissioned by a person who had some special relationship to the events depicted.⁹² Was this the 40+-year-old male whose cremated remains were buried in the Vase? The bones show no definitive evidence for the sharp or blunt force trauma which might be the legacy of an arena career, although due to the severity of taphonomic alteration caused by cremation such trauma cannot be conclusively ruled out. On present evidence it is therefore unlikely that the deceased was one of the performers immortalised on the Vase, but he may have been more than a keen follower of local spectacles, perhaps a lanista (trainer of performers) or an editor (sponsor) of munera.⁹³

If the Vase is interpreted as celebrating a local munus, perhaps the obvious question is, where would this have taken place? Although there is no conclusive evidence for the site of an amphitheatre at Colchester, gladiatorial spectacles were not solely produced within such a venue and could plausibly have taken place in the town's circus or one of its theatres.⁹⁴ There are several known amphitheatres in Britain, but as munera may have been held far less frequently in the provinces, this may in some places have precluded the building of an expensive permanent structure specifically to house gladiatorial games, especially where there were alternatives.⁹⁵ Temporary structures might also have accommodated these spectacles, depending on their scale.⁹⁶ One of Colchester's theatres, in particular, may have been appropriate to housing gladiatorial combat. Its proximity to the Temple of Claudius and its temenos may indicate the sacred relationship between them.⁹⁷ Its role as the first provincial capital meant that Colchester also likely hosted major ceremonies related to the imperial cult – grounds for inferring that gladiatorial munera would have been offered as priestly obligations.⁹⁸ The famous imperial edict of a.d. 176–77, contemporary with the Colchester Vase and intended to mitigate the expenditure games of this kind incurred, speaks to the likely cost of munera in Britain, especially where the provision of performers potentially demanded long and costly journeys.⁹⁹ Colchester's foundation as a veteran colony, a context with which gladiatorial games were especially associated on an imperial level, as well as its potential role within the province, may have instituted a spectacle culture of greater intensity than some other Romano-British cities. The town's favourable location on long-distance communication routes also lent it more privileged access to arena performers, human and animal. This context may help explain the investment in the artistry commissioned to create lasting mementoes of performance such as the Vase.

CONCLUSION

Figured scenes on barbotine-decorated pottery are often referred to as circus or arena scenes and presumed to be generic,¹⁰⁰ but close examination of the figures and inscription on the Colchester Vase points to it having been commissioned from a local pottery workshop to commemorate a record of a specific occasion that took place in the town.¹⁰¹ That it was then used as a cremation urn for a non-local adult male invites further interpretation that he was the commissioner-owner, probably with a closer connection to the event than spectator only. Although the identity of this man remains enigmatic, and although no amphitheatre has been found at Colchester, a holistic reassessment of the Vase has introduced the reality of arena spectacles in the town, ranging alongside or even among the events that would have been held at its circus and in its theatres. The biographical fragments inferred from the inscriptions show how these spectacles drew on Colchester's integration into inter-provincial networks enabling performer mobility and underpinning arena logistics.