Coins contexts and dating

Coins, contexts and dating. Changes are happening in the use of coins in interpreting excavations in the Mediterranean area. The idea that a coin in a context gives a firm date for that context is slowly fading and excavators and interpreters now realise that matters are more complicated than that. What are these complications and should the change spread to the NW of the Empire? I ask the question now because it may well be asked by others in the future and I would like to get my thoughts sorted out before that happens. They are obviously just my thoughts so it also gives others the chance to modify them or refute them as needed. Many of the points made here have been made before by me and by others but it seems sensible to bring them together in one place. First there is not just one date for a coin, there are at least three: the date when the coin was minted, the date when similar coins were commonly in circulation, and the date when that particular coin arrived in the deposit in which it was found. There is worse to follow. There cannot be one set of rules for all Roman coins -- and the same is probably true for coins of any date -- because Roman coins were struck in phases or fashions of shorter or longer duration. During the later part of the republic copper was substantial in form and modest in value which gave it much less reason to travel than the silver which was high in fineness and fairly constant in value. Both attributes will affect the second two forms of date -- common date of circulation and date of deposition. There is less reason for the copper to remain in use/circulation than the silver which is a convenient store of wealth without an obvious end-date. On the other hand the fineness of Republican denarii gives them a limited life in any region under the direct rule of Rome. A system like that of the Empire needs silver to continue to produce silver coins, and the urge to produce more and more will lead to debasement. Old coins are therefore at a premium, as usually containing more silver than new coins -- though this does depend on the actual weight of the two coins being compared -- and it was probably not difficult to remove the older, better coins from circulation while replacing them with newer coins. A modern example is the change in Britain from part-silver to cupro-nickel in 1948 (?), when the banks were charged with removing the old silver coins from circulation. a fact which enabled my father in the bank to bring home every so often shillings and sixpences dating back to George III (c.1816-20). It may be that a careful search will show Republican denarii turning up more often in second, even third century contexts in remote Britain than in close-to-authority Italy. Their minting date is certain within a very short space of time; their date of circulation will depend on the area under study; their date of deposition must depend on context. The system established by Augustus continued through the first and early second century with changes in the silver content of the denarii but there are clear signs of decline towards the year 200. The changes and the signs are not great. For silver they are only now being unravelled in detail, and since the copper depended for its value in its relation to the precious metals, there was no need to reject the smaller later sestertii or save the older larger ones. Where there are peaks in the silver content of denarii, as under Domitian, those finer coins seem fairly quickly to have been removed from circulation. Otherwise coins from Nero onwards have definite minting dates, but a continuous blur of a century or more of circulation dates. Deposition? Using Portchester as an example, there is only one (a first century Claudian Copy) with 602 coins struck after 260. This is in a Shore Fort on the south coast of England which seems to have been founded in the late third century. The coins arrived at Portchester either through use in trade and exchange, just possibly by supply, though I doubt it, but they all date from the later third century onwards. I am unimpressed by the idea that such coins found in contexts of the third and fourth century at London, York, Gloucester or Silchester or other towns show their continued circulation. I prefer to argue from sites where earlier deposits cannot be churned up to provide redeposition. And if the argument is continued I would look to see if any broken pottery of earlier date accompanied those earlier coins, for I do not believe in the continued life of broken pottery without detailed reasoning. The next phase is that of the emperor with a spiked crown of the suns rays -- radiate coins. As a generalisation to which exceptions are not hard to find, radiate coins with striking dates of 250 to 296, form a fairly compact group with a clear start date and a short decline after minting ceased. Where the site being excavated has fairly continuous occupation radiates may well turn up in many later layers, but it would be very unusual to find them in a majority in such contexts. In other words they do have a long tail of residuality. But with a firm date of minting there is also a fairly firm date of majority circulation stretching only a little after the last radiates were struck. The residual tail may extend through the Roman life of the site -- to the seventh century in the Mediterranean. Coinage of the fourth century passed through a series of phases of size, weight, very minor silver addition, and political correctness. To some extent the phases seem to have an end date by which the majority of those issues have left circulation but this also seems to vary from place to place, province to province, and sometimes even site to site. That is, during the fourth century and after, coin use and circulation seem to be a more localised affair than in the earlier empire. This is a snap judgement which needs a future thesis to back it up. It would be nice to compare the coins in twenty substantial fourth century contexts in Britain with the same number in Italy, but if we are dealing with published material that is a vain hope in both areas. Though work at the coal-face with unpublished records from friendly excavators and curators would be possible. It might be possible to use the small number of sites with suspected foundation dates in the fourth century as I used Portchester above. The snag here is that those sites would firstly be mainly military, and secondly dated from written (historical) sources. I have no faith in written sources because they are always written by an individual for a reason -- both of which, in my eyes, invalidate them. The fact that from the other finds, such as equipment, the sites would seem to be military causes further problems because there is the matter of circulation or supply. In the fourth century we are dealing overwhelmingly with copper coinage. For an example, Britain seems richer in silver coins of the later fourth century than the rest of the West and the standard rate of finds here (from excavation, not metal-detecting) seems to be one silver coin for every 1,000 copper coins. Did those copper coins excavated from definite deposits in the site arrive on the site by circulation, exchange, and 'market forces', or were they supplied by the state, to the state servants? We do not know. The silver may well have been supplied, but the copper may well be different. Whatever the reason for their being on the site, they do tend to clump into periods. So a suspected Valentinianic fort tends to have a majority of Valentinianic coins rather than a mixture of earlier material. This needs detailed confirmation. And even then there is the very obvious danger of circular reasoning. The mention of 'political correctness' simply refers to the fact that it is inadvisable to walk round with a purse full of of coins minted by a recently deceased usurping emperor (Carausius, Magnentius) if the thought políce are out in force. Better for those coins to be quite quickly recycled, and typically, they are. In the NW Empire there is an end in view. As usual I am working from British detail supplemented by information from the wider NW. There is a danger here in that I am closely involved with the idea that Roman Britain ended soon after 400. Watch out for circular reasoning -- it did end then because I want it to end then. The latest coins to appear in bulk in Britain were struck from 388 to 395. A smaller number of coins arrived which were struck between 395 and 402, and coins struck after 402 are very rare. They do exist and have been found, but rarely have excavated contexts. There is an obvious point here which prevents any deduction on the lines of 'the coins cease to arrive therefore Romanitas stopped at the same time'. The nearby mints did not produce many, if any coins after the very early 5thc. If they were not minted, they could not arrive and therefore their non-arrival is not evidence for an end. But still, Britain might well have soldiered on with the old coins getting more and more worn and treasured, and that is what the idea used to be. Might it not still be true? In the Mediterranean area coin supply does not stop everywhere around 400. Almost every statement from here on should probably be prefaced with the warning 'on some sites'. There might be a slight dip in production, but issues of the middle of the fifth century are found in quantity and the issues continue till the great change to Byzantine type coins happens around 500; larger flans, typically marked with a numerical value from A (=1) upwards. This is to leave on one side the prolific issues of coinage issued from N Africa after the Vandal conquest. The issues which are commonly found are the smaller values of perhaps 1 to 5. To add the word nummus/nummi to these numbers would be consoling but hardly helpful since we do not know what a nummus (if it existed) represents in market terms. To official Vandal issues there was added a great splurge of copies, sometimes of Vandal issues, and some of those might well be less competent official Vandal issues, but also of almost any issues of the fourth or early fifth century. In Egypt these are small, very thin pieces of struck metal which yet show perfectly recognizable earlier types . At Carthage, just as an example, fourth century coins continue to appear in contexts up to the end of Roman deposits. If the coins from a certain date of minting are listed by the date of the deposit in which they were found then only half the coins from issues after 345 are found in contexts dated before 500. This is worked out in detail with accompanying diagrams in Journal of Archaeological Numismatics vol 2, 265-280, 2012 which is not very often quoted. It can be summed up in the apparently contradictory statements: Many coins of the fourth century were lost at Carthage. Few coins were lost in the fourth century at Carthage. These points, results, and thoughts are being expressed by the excavators and numismatists concerned with other sites in the Mediterranean and it may be reasonable to ask -- Why only now, when the clearance of classical sites has been going on for centuries. The answer has already been hinted at, in that coins from well defined and stratified contexts have only been established and published in the last fifty (?) years. Before that the excavation of the site was described in detail, the architecture studied thoroughly, and the coins well described and identified followed either in an appendix or a separate volume, divorced from the archaeology. Such publications, the rule in any part of the Roman Empire, are of little use for the study of when coins were lost or deposited. And it needs to be added that the coins of the fifth and sixth centuries before the Byzantine reorganisation have only been thoroughly sorted out since the later 20thc. The picture in the Mediterranean area suggests strongly that coins minted in the 4thc continued to be deposited (and therefore in use?) well into the sixth century. Should this information not be fed back to the NW provinces so that, for example, Roman Britain may be freed from the chains of 400 and so continue using 4thc coins, without the addition of new issues, for many years after? The problem has been an essential part of the discussion so far. Stratigraphy. Contexts. If Roman Britain is to be allowed to continue past the first occurrence of coins of 388-402 on the basis that 4thc coins can circulate perfectly well for decades after their minting, there is the need for contexts, deposits, in which they can be found. While layer after layer is added to the stratigraphy of Carthage, Rome, Butrint, Beirut, on top of coins of c.400 the same simply does not happen in Britain except in sites which later became medieval towns. The villa, farmstead, hut or temple typically has a stratigraphy which goes up to layers with coins of the fourth century and then simply stops. In town sites there are later layers on top of Roman layers but there is usually a clear discontinuity of deposit, or gap of sterile material. When deposition picks up any Roman coins are typically a disorganised mixture of any of the coins found in the earlier Roman layers, and their composition does not change from layer to later layer as the residual coins continue to be recycled. On this occasion we cannot join the excellent new ideas from abroad. Richard Reece