Orton C., Tyers P., Vince A. Pottery in archaeology

Подождите немного. Документ загружается.

Implications for practice 35

you have?' (usually in percentage terms) or even 'Do you have any assem-

blages like this one?' To answer such questions, we need a reliable method of

quantification. Archiving in general is discussed in chapter 8 and quantifi-

cation in chapter 13.

The rapid increase in the volume of excavated material, especially in

Britain, has raised questions about disposal versus retention, and of sampling

as a way of reducing the volume of material retained. There are no easy

answers, but forewarned is forearmed. Our views are expressed in chapter 8.

Finally, there comes publication, which is only the tip of an iceberg, the

bulk of which is the archive and the retained material itself. The aims of

publications vary greatly, depending on the nature of the site and its pottery,

the amount and nature of previous work done in the area, the existence of

relevant type-series and the opportunity (or lack of it) to raise one's nose

from the grindstone and synthesise a little. But at the very least, the publi-

cation should act as a set of signposts to its archive and through it to the

objects themselves, so that readers can tell whether they need to consult the

archive and/or examine any of the artefacts.

Some archaeologists find publication psychologically difficult: it has an

awful finality not unlike that of death. Part of the dread comes from the

mistaken belief that what we say will be the last word on the subject, so it had

better be 'right'. But in archaeology there are no last words, all is provisional,

and if no-one ever improves on our work it is not because it is perfect but

more likely because it is terminally boring. Approach publication in this

spirit, and with the advice of chapter 9 in mind, and it may not seem quite

such a burden.

PART II

PRACTICALITIES: A GUIDE TO

POTTERY PROCESSING AND

RECORDING

INTEGRATION WITH RESEARCH

DESIGNS

The archaeologist is in an unenviable position every time he or she plans a

field project, since archaeological fieldwork destroys the subject of its study.

This is true of a field survey almost as much as it is of an excavation, unless

artefacts are examined where found on the ground and left in place. Almost

all people working on sites previously investigated by an earlier generation of

archaeologists have wished at some time that their predecessors had taken up

some other profession. Those involved in an archaeological project therefore

have a responsibility not only to carry out their stated aims but also to

integrate their work into that of their predecessors and to ensure that their

methods of recovery, analysis and recording are going to produce a usable

archive for future workers. In some parts of the British Isles, for example, it is

estimated that modern development, mineral extraction and agricultural

practices will have brought field archaeology to an end within a few decades.

this

should prove to be the case then those lucky enough to be involved in

fieldwork now will be creating the only research materials available to their

successors.

Nevertheless, the most important duty of any pottery researcher is to

ensure that the recovery, analysis and recording of the pottery from a project

is carried out smoothly and efficiently and within previously agreed limits of

time and money. The practical steps should be: (i) to estimate the likely

quantity of pottery which might be recovered during the project; (ii) to read

and absorb previous work in the study area and, where possible, to build

upon this work; (iii) to be aware of the best practice in the field and to adopt

it, unless compelling arguments can be made for not doing so; and (iv) to

produce and cost a strategy for allowing the estimated volume of pottery to

dealt with as part of an overall research design for the project (Fulford and

Peacock 1984).

The archaeologist in charge of a project may have little practical experience

of pottery work and a limited series of objectives, which if carried through

might produce only a fraction of the potential information from the pottery.

Where the site or area under investigation was pottery-using it is common to

find that pottery, along perhaps with animal bones, forms the bulk of the

finds and potentially the bulk of the expenditure on the project. In these

circumstances the project manager needs good advice and the strategy which

is adopted must be talked through by all concerned so that as many of the

40 Integration with research designs

potential pitfalls as possible can be foreseen and avoided. In the end, of

course, there is no substitute for experience and it is too much to expect

anyone with a general training in archaeology to immediately step in and run

a large pottery project.

If it is envisaged that the pottery will be eventually studied by specialists

who are not going to be part of the actual field team, then these people should

be consulted before the start of the project. There may be special recovery or

recording requirements about which the field team should know. Further-

more, the way to get the most from a specialist is to make sure that he or she

feels

part of the project and believes that their ideas and knowledge have been

properly used.

The standards and methods used by pottery researchers have changed so

quickly during the twentieth century that there is a temptation to assume that

nothing of value is to be found in previous work. This would be quite wrong.

Apart from anything else, people working in the early to mid twentieth

century quite often worked on a much larger scale than can be contemplated

today and had first hand practical experience. If nothing else, one can

sometimes spot the origin of particular ideas and evaluate their worth rather

than having to accept them without question or going back to first principles

and establishing every point from scratch. This is a costly, and wasteful,

process.

Because of the slow speed at which archaeological work is published one

may well run up against the problem that crucial work has been completed

and written up but is not yet in print. There is no satisfactory solution to this

problem. It might be possible to see a preprint of the paper but it can hardly

be properly evaluated without the accompanying site evidence. The solution

for the future is to prepare and deposit in a public archive a report, here

referred to as an archive report (see p. 98), which contains a digest of the

results of your work on a particular collection. Previous work was usually

organised on the assumption that the finished report was the final word and

that most of the paraphernalia, notes, record cards or illustrations, were

ephemeral, being of value to the author but perhaps misleading in anyone

else's hands.

At the very least, one should track down and examine any collections of

pottery previously recovered from the vicinity of the planned fieldwork. Even

without an accompanying report or stratigraphic data, such collections give

you an idea of the range of forms and fabrics to be expected. If the work has

not already been done, you can use these vessels as the basis for your fabric

and form series (see p. 72 and p. 77).

Another task which should be done before the project starts is to look at

what is known about pottery production and clay sources in the study area.

Geological memoirs often include a section on economic geology which will

list those beds of clay which are known to have been used for pottery, brick or

41 Integration with research designs

tile manufacture. In some parts of the world traditional methods of manufac-

ture continue to be carried out and it is possible to find out at first hand what

resources and production methods are being used. Finally, one should know

as much as possible about the main export wares of the period and area. In a

region where little pottery work has been undertaken previously these wares

may form your first means of producing some sort of local chronology.

The most important function of any strategy for dealing with the pottery

from a project is to allow the aims of the project to be fulfilled. They will

almost certainly include the dating of activity on a site and the broad period

of occupation of sites and may include comparison of activity in different

areas or sites, looking for evidence for or against social differentiation or

specialised activities, for example.

To achieve these aims it may also be necessary to examine aspects of the

archaeological record which are not of primary interest. For example, the

processes by which the archaeological record came into existence must be

understood before any further use can be made of the data (see p. 209).

There will almost certainly be other aspects of the pottery data which could

be examined but which are neither of primary interest nor crucial to further

interpretation. The temptation is therefore not to record these aspects at all.

For example, there is certainly information to be found in studying pottery

assemblages deposited in similar conditions and circumstances on different

sites. A good example is the post-medieval cesspit assemblage. Pottery has

been collected from backfilled cesspits in both Europe and the Americas and

comparison of one group with another should reveal both similarities and

divergences in culture on either side of the Atlantic. The work involved in

examining even the material from a single pit is enormous and it is important,

if any information is ever to be extracted from this data, that there is some

standardisation in the way in which the contents are recorded. There is a

similar interest in standardisation from those concerned with the reconstruc-

tion of marketing patterns and trade routes through the study of pottery (see

p. 197). Studies, such as those of Hodder and Fulford on the distribution of

Romano-British pottery used data from existing collections and managed to

find some interesting patterns within them (Hodder 1974; Fulford and Hodder

1974). Further advances, however, require the study of larger collections,

with some sort of stratigraphic control, and it would be wasteful, if indeed it

were possible, for an individual to record or re-record all the necessary data

him or herself. This type of study too requires a measure of cooperation

between researchers.

On many excavations it has been the practice to treat certain classes of

pottery in a different way from that of the mass of pottery. Examples are very

common in the field of Romano-British pottery studies, where for many years

samian ware, mortaria and amphoras have been studied by separate special-

ists, so that it has not been possible to study the relative frequency of these

42 Integration with research designs

wares as a proportion of the total ceramic assemblage on a site (Fulford and

Huddleston 1991,9-11,48).

Data therefore need to be recorded in such a way that they can be used by

other workers. The most obvious area of standardisation concerns termin-

ology and classification, although the names used for a particular class are

irrelevant so long as it is possible to translate from one worker's classification

to another's. Quantification is a less obvious area for standardisation and

yet

all examples given above require standardised methods of quantification (see

p. 166).

Having emphasised the need for cooperation and standardisation, the case

for ensuring some sort of continuity within a region must now be made. If

there is a system of classification and recording in operation within a region

then any move to abandon this system must be thought through carefully.

Change for the sake of change is pointless and will have the effect of making

all previous records less accessible and less useful. Such is the speed of change

in pottery studies that there are excavation units in the United Kingdom

which have been in existence for less than two decades but which nevertheless

have three or more incompatible systems of recording amongst their records.

To use computer jargon, it is important to ensure upwards compatibility

when a system is modified. At Lincoln and London many data sets have been

transformed from earlier systems into current formats. It is never as simple as

it might appear and usually requires re-examination of the potsherds them-

selves. This creates problems of logistics, especially if the material is stored in

a different place from that where the work will be undertaken.

Archaeology does not take place in an ivory tower and the factors which

eventually lead to the choice of system to be used on a project will include

time and money as much as the aims of the project, the potential uses of the

data and the need to work within an evolving discipline. Potsherds can be

classified and recorded in the most minute detail and some of the recording

systems developed and used in England in the 1960s and 1970s recorded

virtually every characteristic which could be recognised, from fabric and

form, through sherd thickness and hardness to the colour of the core and

surfaces of the sherd. Fashion has moved on, and it is currently in vogue to

record the bare minimum as a basic record. It may seem obvious that the

more one records the longer the recording will take, but there is always a

temptation to record in more and more detail and a tendency to forget that

while it may only take a few seconds longer to record two traits instead of

one, on a single occasion a typical site pottery collection might contain tens

thousands of sherds. The more common a trait is, the greater the possibility

that its study might be rewarding in terms of revealing a pattern but the more

time and effort will be added to the project. By contrast, rare features, such

complete, highly decorated vessels, might be recorded in tremendous detail

without adding appreciably to the length of the project, but without adding

43 Integration with research designs

much to the sum of knowledge either. As in life in general, you don't get

anything for nothing.

To summarise, in this section we have looked at the importance of project

design and of making sure that the system of pottery study adopted will not

hinder any of the aims of the project being

fulfilled.

We then looked at the

need to collect data for others, in the expectation that they might do the same

for you one day, and the importance of keeping some form of compatibility

between the classification and recording systems adopted for the future and

those used in the past. Finally, we looked at the importance of time and

money in the design of pottery recording systems. From all of this we can

conclude that there is no single, right method of recording pottery from an

archaeological project and that every practical system will be a compromise

between conflicting demands.

LIFE IN THE POT SHED

Introduction

In this section we look at the stages by which pottery is collected from an

excavation or field project, and how it is identified, recorded and analysed

before, eventually, being consigned to storage in a museum or archaeological

resource centre. The first stage is to set up a base but this is quickly followed

by the adoption of a retrieval strategy - to sieve or not to sieve. There are

occasions where one deliberately does not collect all the pottery present in a

deposit and the rules governing collection policy are discussed next. We then

look at the way in which finds are processed and the initial sorting of pottery

from other finds. At this stage it is common for there to be a preliminary

viewing of the pottery, referred to here as spot-dating. There may, however,

be a considerable interval after this point before 'serious' analysis of the

collection begins. We start by looking at the typical flow of work, from the

laying out of the collection to its identification and recording, and then look

at the range of further analyses and processes which may be required for

specific sherds. Then follows a digression while we look at the various

exceptions to these general rules (classes of ceramic artefacts which need to be

treated differently). They might, like building materials, need to be recorded

in less detail than potsherds or, like, kiln waste, they may lend themselves to a

series of special questions or, like crucibles, they may relate directly to

non-ceramic finds. Finally, we look at the structure of the pottery records, the

use of computers in data storage and analysis and the integration of pottery

records with other databases.

Setting up base

The location of the finds processing for a project will depend very much on

the availability of work space and resources. On a traditional rural exca-

vation the base might be a tent or caravan whilst in an urban excavation parts

of the building being demolished might be available for use. Large excavation

units with several excavations running concurrently may wish to centralise

finds processing so as to make economies of scale and to enable workers to

move between material from one site and another as the work flow varies. In

any particular case there may be no option but to take whatever is available

and adapt it as one can. However, there are certain rules and requirements

which can be set out and used wherever choice exists.

Setting up base

In hot dry climates the provision of water for washing and desalination of

finds may be a problem. It is not uncommon for all the water needed by a

project to have to be imported. Where salts encrust the finds they must be

soaked in fresh water for a considerable time. Muslin or netting bags can be

used to separate the material from several contexts in a single container. In

the United Kingdom the more normal problem is not the provision of water

but its disposal. Normal domestic drains, for example, will rapidly become

clogged with silt and so silt traps must be added below the sink. Periodically

they will require emptying and so there must be some means of disposing of

quantities of wet sludge.

Adequate lighting is also important. People will be sitting down washing

the edges of sherds and marking their surfaces. They need to see what they are

doing. Furthermore, if the pottery is to be spot-dated on site then access to

direct sunlight is an advantage. If this is impossible then a desk or a table

lamp per person or some similar light source will be needed.

On a productive excavation there may be very large quantities of finds

being processed. They can be washed much more quickly than they can be

dried and there will come a time when every available space has been covered

with trays of drying pottery, bone and so on. Racking and mesh-bottomed

trays can alleviate this problem, but it is very easy to overlook it, especially if

one is looking at a potential work area as being for a single person. Whenever

conditions become crowded, and in particular where the pot shed is part of a

general work area, or forms part of a means of access, there is the danger that

someone will trip over trays of drying finds and the material from more than

one context will become mixed. Another danger of making the pot shed too

accessible is that people wander through, pick up sherds and put them back in

the wrong tray. No amount of warning or scolding will do any good -

especially if the culprit is the site director or a distinguished visitor - and it is

better to make sure the problem does not arise by keeping the drying areas at

some distance from the other work areas.

If the pot shed is also to be used for the post-excavation work then a large

amount of bench space will be needed. The ideal work space for the material

from a large excavation would be the size of an aircraft hangar. Few urban

projects are able to supply suitable working areas and consequently their

work tends to suffer.

It may well be possible to provide much better accommodation at a

distance from the excavation but this will also bring problems. Firstly, the

finds will have to be transported from the site to the pot shed. To ensure that

all the finds arrive safely they should be listed at one end and checked off at

the other. Secondly, the further away from the site the pot shed is physically

the further away it is socially. There was an occasion where atrocious weather

led to the calling-off of excavation for the day and the team headed into the

nearest town in their Landrover. It was only when they arrived that someone

Life in the pot shed

remembered the finds supervisor, who was obliviously washing pottery

in Jjj

outhouse at the local farm. This does nothing to encourage a team spirit.

Should the option of centralising finds processing be taken then these

problems are exacerbated. A division into 'them' and 'us' will arise both on

site and in the pot shed and this will lead to difficulties for the finds staff jj

monitoring site recovery and collection and the suspicion by the site staff that

the money could have been better spent on more excavation. Wherever

possible it is much better to have a finds specialist at hand on site. They can

give instant advice and identification and be able to adapt recovery and

processing to the needs of the site.

Collection policy and practice

Much of the information potential of your pottery will depend on the

methods used to collect it and the rigour with which they are applied.

Experiments using different methods of recovery and excavation have

emphasised that recovery methods not only control how much of the total

pottery in a deposit you recover but can lead to biases in the data.

To take the most brutal form of excavation, using a pick and shovel, as an

example, only the largest sherds will be seen by the digger and this is bound

create a bias against the recovery of sherds from vessels which were either

not

very big when complete or tend to fragment into smaller pieces, perhaps

because of the thinness of their walls or their brittleness. Hand recovery using

a trowel or similar tool can be very thorough, depending on the speed with

which it is carried out, the contrast in colour between potsherds and the soil

matrix and the consistency of that matrix. Clayey soils may break up into

large lumps which might contain potsherds whereas sandy soils disintegrate

entirely into their constituent grains, making it much easier to see finds. There

are various ways in which these biases can be corrected and, indeed, it may

well be that having recognised that the bias exists there is no need to correct

it. One method of correction might be to excavate a sample of a deposit using

a trowel and then remove the rest by pick and shovel. One can make a case

for

saying that if one is interested in the relative proportions of different types

and the range of types present in an assemblage of pottery then it is much

better to have a medium-sized, unbiased sample than to have a large but

biased one.

The use of sieves to screen the soil is essential for the standardised recovery

of objects whose maximum dimension is less than c. 30mm. This is not to say

that smaller objects will not be retrieved by hand, only that they will be found

less frequently than they should be. Different methods of sieving have been

developed, depending on the objectives of the exercise and the consistency

the soil. On clayey, silty or organic deposits it is essential to use water to

break down the lumps. Chemicals such as hydrogen peroxide will make this

process more efficient, but these are expensive to use on a large scale. It