'Celtic Fields' – Archaeology's stepchildren
Traces of prehistoric farming in Western, Central, Eastern and Northern Europe:
Sondages in ‚Celtic Field‘ rims to get dating material (back to Finding 'Celtic Fields')

The following explanations and recommendations are based on the author's personal experience, which was gained mainly in northern Germany and may not necessarily be transferable to other areas, especially those further south. The search for datable material from archaeological sources generally presupposes that livestock were already kept indoors and that the plots were at least occasionally fertilised with the dung produced there, which is composed of stable litter and manure and in which domestic waste can also be found. In northern German ‚Celtic Fields‘, preserved remains of domestic waste in the form of charcoal and very small shattered pottery are usually found in well-developed field margins. In addition, there is usually also annealed flint, which is relatively easy for the trained eye to recognise. While the pieces of pottery, due to their small size, usually only provide approximate dating information, the charcoal can provide more precise C14 or radiocarbon dates, especially if short-lived material such as charred seed grains, hazelnut shells or pieces of small twigs are used. However, the quality of the dating depends crucially on the steep ness of the calibration curve. In this context, the "Hallstatt plateau", a long period in the middle of the first pre-Christian millennium in which charcoals of different ages have more or less the same C14 age, is particularly feared.
Of course, there are also areas largely without domestic waste, which may have been far away from the associated settlement and therefore "got nothing". There, it might be possible to find charcoal that originates from initial fire clearance or from the burning of cleared wood that spread over the meadows or hedges that may have been planted there.

My very first three sondages in southern Germany (south of Frankfurt/Main) did not reveal any domestic waste except for a tiny ceramic potsherd and very little charcoal. Whether this is to be taken as an indication that no manure from stables was brought there must remain open, as far as no further evidence for this is available yet.
The first step is to find the most favourable locations for sondages. As a rule of thumb, it has been confirmed that sondages are most productive where arable farming was most intensive and/or of longest duration, i.e. where the most prominent marginal heights have formed. In any case, one should avoid areas where larger quantities of modern charcoal are to be expected, be it the remains of burnt wood waste after clearing or the remains of charcoal production in kilns. With some experience you can recognise both quite well: Charcoal of modern burnt wood waste is light and does not show any contamination of the outer vessels under the stereomicroscope, and charcoal from kilns is rather heavy (if not failed), because the charcoal burners always tried to control the smouldering process in such a way that as many flue gas components as possible condensed in the charcoal and thus increased its weight and calorific value. Accordingly, the smaller wooden vessels are often dense under the microscope. By the way, the areas directly surrounding visible ground monuments such as burial mounds should also be avoided, not only for reasons of heritage conservation, but because activities may have been carried out there that led to an accumulation of charcoal which has nothing to do with the traces of arable farming.
The main intention of the sondages is to sieve (almost) everything. Therefore, the sondages only need to be small in size, usually about 60 x 85 cm. Only the uppermost 5-10 cm are removed unscreened, in the case of a thick layer of raw humus, e.g. under conifers, possibly a little more. All the rest is sieved in separate removal layers of 5-10 cm, as dry as possible. Ideal, but not always applicable, are sieves with a mesh size of 2 mm and a surface of at least 40 x 60 cm; if necessary, only a part is sieved in this way and the rest with double the mesh size. A high tripod with holding chains for the sieves is recommended. All larger stones should be removed from the sieve immediately, as they may crush the sensitive charcoal, but the stones should be checked carefully for artificial traces.
If the soil is too cohesive, dry sieving is not possible. Then wet sieving must be done with specially designed bucket sieves that are immersed in tubs filled with water. This generally works with a sieve with a 2 mm mesh size. If possible, the sieve residues are afterwards pre-cleaned with a battery-operated sprayer and roughly screened in the sieve. It is advisable to remove as much water as possible from the sieve residues while they are still in the sieve. Dabbing with absorbent old rags from below has proven to be a good method for this. It is best to wrap the screen residues in thick layers of absorbent paper to facilitate subsequent drying. The main problem here is often getting the large amount of water needed for this.

The highly simplified scheme is intended to illustrate why a search in very slightly elevated field rims is hardly useful. Of course, many other factors also affect the preservation of charcoal and ceramics, especially bioturbation and the associated changes in position.

The necessary depth of the sondage is often difficult to estimate. If the average height of the embankment above the roughly level surroundings is about 20 cm (which is already a lot), archaeological finds are unlikely to reach deeper than 40 cm, unless one happens to come across a deepened archaeological feature. In terraced areas, these values can be much higher, and the depths of the sondages should readily exceed 60 cm.
In any case, at least one of the sidewalls of the sondage should be documented after soil removal. By the way: to recognise the former surface in the form of a buried humus band is rather unlikely. In general, the deep browning of our forest soils has obliterated these traces, if the old surface was ever preserved. There is a greater chance of this happening in sandy, highly acidified podsol soils. This is why it is often difficult to determine the boundary between ploughed-through soil and the "grown soil" underneath. After a good photographic and/or drawing documentation, there is now the opportunity for further sampling in the profile, e.g. for various pedological investigations, archaeobotany, phytolith determination and in deeper accumulation layers also for optically stimulated luminescence (OSL) as an alternative age determination. It is good, but rarely feasible, if experienced soil scientists examine the profile on site.

Experience shows that it is only possible to look through the sieve residues very superficially on site, which is, however, quite useful, e.g. for determining the maximum depth of the sondage. Especially in the forest, the root penetration is so intensive and deep that thousands of roots in the sieve residues make screening difficult, so it helps a little to sort out the coarsest ones during the sieving process. Normally, even with dry sieving, it is necessary to wash the residues again, for which a battery-powered sprayer is well suited. The subsequent washing, drying on large trays and intensive screening of the sieving residues should be done indoors, or alternatively in a shelter protected from the wind (charcoal and finding notes blow away easily!). The residues should be spread out on absorbent paper to dry. The most favourable stage for finding the charcoal is shortly before it is completely dry, when the charcoal, which holds the water particularly well, stands out from the other particles by its still deep black colour. The thorough examination of the residues generally requires more time than the construction of the sondage, as it is necessary to work with magnifying glasses in order to also find charred seed grains or remains of charred nutshells, which are particularly suitable for age determination. Knowledge of the local spectrum of prehistoric pottery is of course useful, and it requires some practice to recognise even very small fragments of this pottery.
As an alternative to sondages, it is also possible to search the root plates of recently fallen trees if they stood exactly on the edges of a field and the soil is cohesive enough for the roots to hold it. The depth of the sediment to be sampled and sieved can usually be estimated to some extent. The disadvantage of this method, apart from an expected greater inaccuracy in determining the depth of removal, is above all that there are many more roots and rotten wood in the sieve residues, plus an unexpected habitat of many insects of all stages, especially during the period of winter dormancy. This method is rather recommended for first, preliminary dating.
At bigger sondage depths, e.g. in very pronounced plot edges or very distinct terraces, one must reckon with a time of origin of several hundred years for these edges, which could also manifest itself in the dating of the uppermost and lowermost charcoal samples. In the Riesewohld in Dithmarschen, for example, a period of use of about 600 years can be assumed on the basis of two dated charred hazelnut shells at the very bottom and at the very top of the profile. Apart from that, there are many possibilities for misdating. Older charcoal, e.g. from earlier epochs such as the end of the Ice Age or the early post-glacial period, may have found its way into the plough soil, but younger charcoal may also have found its way into older layers through bioturbation. Therefore, as much data as possible should always be collected, which of course often fails due to financial possibilities.
Frequent finds are small, round and hollow seemingly charred globules of 1 - 3 mm diameter: sclerotia of the ectomycorrhiza fungus Cenococcum geophilum, which forms symbiotic communities with trees especially in old-growth forests. These sclerotia can, but do not have to be particularly old, partly they are still capable of germination and are therefore not suitable for dating.
Of course, sondages require permission, not only from the forest owner. Here I could sing praises about German bureaucracy. While in some cases permission was granted amazingly unbureaucratically and without many conditions, the lower heritage conservation authority in Burgenland in Saxony-Anhalt sent me, after half a year of waiting, a 14-page form that had to be submitted in triplicate with several attachments. So far, I have gratefully dispensed with this 50-page stack of paper. I had a similar experience in Dreieich near Frankfurt, where the archaeological heritage department agreed without any problems, but HessenForst insisted on obtaining permission from the lower nature protection authority and above all from the explosive ordnance clearing service, which made (expensive) external investigations a condition. Here, too, sondages had to be avoided for reasons of cost, but the Frankurt municipal forestry administration issued permits for its neighbouting area without such conditions, so that some sondages could be carried out after all. – Back to Finding 'Celtic Fields'

Back © Volker Arnold 2022